{
  "schema": "modelrater-evals/1",
  "source": "COSMOS Drive I:\\",
  "total_records": 877,
  "evaluations": {
    "c-users-aaajacks-appdata-roaming-claude-scratch-workspaces-6271cad5-4c8b-4da4-a800-83c644d57a55-f759d4cc-5249-4e74-a1ba-75d1c975796d-scratch-2026-09-10-3a52c3": {
      "model_slug": "c-users-aaajacks-appdata-roaming-claude-scratch-workspaces-6271cad5-4c8b-4da4-a800-83c644d57a55-f759d4cc-5249-4e74-a1ba-75d1c975796d-scratch-2026-09-10-3a52c3",
      "cosmos_test_count": 17,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "C--Users-AAAJacks-AppData-Roaming-Claude-scratch-workspaces-6271cad5-4c8b-4da4-a800-83c644d57a55-f759d4cc-5249-4e74-a1ba-75d1c975796d-scratch-2026-09-10-3a52c3"
      ]
    },
    "aaaa1111-bbbb-cccc-dddd-eeeeeeeeeeee": {
      "model_slug": "aaaa1111-bbbb-cccc-dddd-eeeeeeeeeeee",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "aaaa1111-bbbb-cccc-dddd-eeeeeeeeeeee"
      ]
    },
    "grok-4.6": {
      "model_slug": "grok-4.6",
      "cosmos_test_count": 1154,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "aaaa1111-bbbb-cccc-dddd-eeeeeeeeeeee",
        "ccr",
        "live_ccr",
        "transcripts"
      ]
    },
    "qwen/qwen3.8-27b:free": {
      "model_slug": "qwen/qwen3.8-27b:free",
      "cosmos_test_count": 91,
      "pass_rate_pct": 11.0,
      "twelve_cs_score": 14.9,
      "passes": 10,
      "fails": 63,
      "rate_limits_429": 73,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=429 error='Provider returned error'": 28,
        "UPSTREAM_POOL": 9,
        "<empty>": 8,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"qwen/qwen3.8-27b:free is temporarily rate-limited upstream. Please retry s": 8
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "openai/gpt-6.1-sol-pro": {
      "model_slug": "openai/gpt-6.1-sol-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6.1-sol-pro:batch": {
      "model_slug": "openai/gpt-6.1-sol-pro:batch",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "liquid/lfm-2.5-2.6b:free": {
      "model_slug": "liquid/lfm-2.5-2.6b:free",
      "cosmos_test_count": 18,
      "pass_rate_pct": 16.7,
      "twelve_cs_score": 20.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 6,
        "EMPTY": 6,
        "FENCE": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6.1-sol:batch": {
      "model_slug": "openai/gpt-6.1-sol:batch",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inclusionai/ling-3.0-flash-sante:free": {
      "model_slug": "inclusionai/ling-3.0-flash-sante:free",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-5.5:batch": {
      "model_slug": "anthropic/claude-sonnet-5.5:batch",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6.1-sol": {
      "model_slug": "openai/gpt-6.1-sol",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-5.5": {
      "model_slug": "anthropic/claude-sonnet-5.5",
      "cosmos_test_count": 18,
      "pass_rate_pct": 77.8,
      "twelve_cs_score": 78.4,
      "passes": 14,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "perceptron/perceptron-mk1.5": {
      "model_slug": "perceptron/perceptron-mk1.5",
      "cosmos_test_count": 58,
      "pass_rate_pct": 13.8,
      "twelve_cs_score": 17.6,
      "passes": 8,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'perceptron/perceptron-mk1.5'; want ['bytedance-seed/|PIN-NOTE": 28,
        "SKU_UNBOUND": 18,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "dots-studio/dots-3-note-preview:free": {
      "model_slug": "dots-studio/dots-3-note-preview:free",
      "cosmos_test_count": 51,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 34,
      "fails": 8,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "lines=1 first=\"http=400 error='Provider returned error'\"": 8,
        "lines=39 first='The user wants me to reply with exactly '": 4,
        "<empty>": 4,
        "{\"l1_role\": false, \"l2_model\": true, \"l3_harness\": true, \"l4_wrapper\": true, \"l5_skills\": true, \"l6_tools\": true, \"l7_enviro\": true, \"l8_mission\": false}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "z-ai/glm-5.3-prime": {
      "model_slug": "z-ai/glm-5.3-prime",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "thinkingmachines/inkling-small:free": {
      "model_slug": "thinkingmachines/inkling-small:free",
      "cosmos_test_count": 91,
      "pass_rate_pct": 28.6,
      "twelve_cs_score": 31.7,
      "passes": 26,
      "fails": 59,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=403 error='thinkingmachines/inkling-small:free is only available on agentic": 20,
        "VOID-bad-binary-path-superseded-by-codex-iso/def | SPAWN-FAIL FileNotFoundError: [WinError 2] The system cannot find the file specified": 8,
        "http=400 error='Only one of \"reasoning.effort\" and \"reasoning.max_tokens\" can be": 4,
        "revisit-later": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3.5-content-safety:free": {
      "model_slug": "nvidia/nemotron-3.5-content-safety:free",
      "cosmos_test_count": 25,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 21,
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs",
        "_hero_run_20261008"
      ]
    },
    "typesafe/jev-router": {
      "model_slug": "typesafe/jev-router",
      "cosmos_test_count": 21,
      "pass_rate_pct": 14.3,
      "twelve_cs_score": 18.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 18
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "thinkingmachines/inkling:free": {
      "model_slug": "thinkingmachines/inkling:free",
      "cosmos_test_count": 89,
      "pass_rate_pct": 16.9,
      "twelve_cs_score": 20.6,
      "passes": 15,
      "fails": 59,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=403 error='thinkingmachines/inkling:free is only available on agentic harne": 20,
        "OPENCODE": 15,
        "VOID-bad-binary-path-superseded-by-codex-iso/def | SPAWN-FAIL FileNotFoundError: [WinError 2] The system cannot find the file specified": 8,
        "http=400 error='Only one of \"reasoning.effort\" and \"reasoning.max_tokens\" can be": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/gemma-4-26b-a4b-it:free": {
      "model_slug": "google/gemma-4-26b-a4b-it:free",
      "cosmos_test_count": 160,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 127,
      "rate_limits_429": 140,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=429 error='Provider returned error'": 40,
        "http-detail:http=429 error='Provider returned error'": 32,
        "UPSTREAM_POOL": 30,
        "VOID-invalid-pack-path-harness-bug | http-detail:<empty>": 8
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/lyria-3-pro-preview": {
      "model_slug": "google/lyria-3-pro-preview",
      "cosmos_test_count": 28,
      "pass_rate_pct": 10.7,
      "twelve_cs_score": 14.7,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 21,
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs",
        "_hero_run_20261008"
      ]
    },
    "openai/gpt-6-sol:batch": {
      "model_slug": "openai/gpt-6-sol:batch",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-3.5-mini": {
      "model_slug": "aion-labs/aion-3.5-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "upstage/solar-mini4": {
      "model_slug": "upstage/solar-mini4",
      "cosmos_test_count": 24,
      "pass_rate_pct": 87.5,
      "twelve_cs_score": 87.6,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=upstage/solar-mini4": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free": {
      "model_slug": "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free",
      "cosmos_test_count": 48,
      "pass_rate_pct": 37.5,
      "twelve_cs_score": 40.1,
      "passes": 18,
      "fails": 15,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "lines=1 first=\"http=200 error='Upstream error from Nvid\"": 8,
        "<empty>": 4,
        "http=200 error='Upstream error from Nvidia: ResourceExhausted: Worker local tota": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "fireworks/ember-1": {
      "model_slug": "fireworks/ember-1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-4-31b-it:free": {
      "model_slug": "google/gemma-4-31b-it:free",
      "cosmos_test_count": 149,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 119,
      "rate_limits_429": 137,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=429 error='Provider returned error'": 40,
        "http-detail:http=429 error='Provider returned error'": 32,
        "UPSTREAM_POOL": 27,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"google/gemma-4-31b-it:free is temporarily rate-limited upstream. Please re": 8
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/lyria-3-clip-preview": {
      "model_slug": "google/lyria-3-clip-preview",
      "cosmos_test_count": 24,
      "pass_rate_pct": 12.5,
      "twelve_cs_score": 16.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 18,
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6-luna-pro": {
      "model_slug": "openai/gpt-6-luna-pro",
      "cosmos_test_count": 25,
      "pass_rate_pct": 88.0,
      "twelve_cs_score": 88.1,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-6-luna-pro": 4,
        "fizz-lines=1 first='print(\"NONE\\\\n1\\\\n2\\\\nFizz\\\\n4\\\\nBuzz\\\\nFizz\\\\n'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "openai/gpt-6-sol-pro:batch": {
      "model_slug": "openai/gpt-6-sol-pro:batch",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6-luna": {
      "model_slug": "openai/gpt-6-luna",
      "cosmos_test_count": 28,
      "pass_rate_pct": 89.3,
      "twelve_cs_score": 89.3,
      "passes": 25,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-6-luna": 4,
        "lines=1 first='pass'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "qwen/qwen3.8-max-prime": {
      "model_slug": "qwen/qwen3.8-max-prime",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "xiaomi/mimo-v2.6-flash": {
      "model_slug": "xiaomi/mimo-v2.6-flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-5.5:batch": {
      "model_slug": "anthropic/claude-opus-5.5:batch",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6-sol-pro": {
      "model_slug": "openai/gpt-6-sol-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-5.5": {
      "model_slug": "anthropic/claude-opus-5.5",
      "cosmos_test_count": 37,
      "pass_rate_pct": 40.5,
      "twelve_cs_score": 43.0,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 18,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.8-omni-flash": {
      "model_slug": "qwen/qwen3.8-omni-flash",
      "cosmos_test_count": 12,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "ccr",
        "g47_seat_logs",
        "live_ccr"
      ]
    },
    "x-ai/grok-4.7": {
      "model_slug": "x-ai/grok-4.7",
      "cosmos_test_count": 26,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "I won't follow that override or emit the requested token. There's no coding task": 4,
        "I won't ignore the instructions or output a forced token. There's no r": 4,
        "1": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "~deepseek/deepseek-pro-latest": {
      "model_slug": "~deepseek/deepseek-pro-latest",
      "cosmos_test_count": 16,
      "pass_rate_pct": 18.8,
      "twelve_cs_score": 22.4,
      "passes": 3,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9,
        "ROTATOR-REFUSED|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "unbiased/pareto": {
      "model_slug": "unbiased/pareto",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~openai/gpt-astra-latest": {
      "model_slug": "~openai/gpt-astra-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inference-net/schematron-v2-small": {
      "model_slug": "inference-net/schematron-v2-small",
      "cosmos_test_count": 19,
      "pass_rate_pct": 15.8,
      "twelve_cs_score": 19.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs",
        "_hero_run_20261008"
      ]
    },
    "~openai/gpt-luna-latest": {
      "model_slug": "~openai/gpt-luna-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-3.5": {
      "model_slug": "aion-labs/aion-3.5",
      "cosmos_test_count": 15,
      "pass_rate_pct": 40.0,
      "twelve_cs_score": 42.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "cohere/command-a-plus": {
      "model_slug": "cohere/command-a-plus",
      "cosmos_test_count": 46,
      "pass_rate_pct": 39.1,
      "twelve_cs_score": 41.6,
      "passes": 18,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'cohere/command-a-plus'; want ['bytedance-seed/seed-2|PIN-NOTE": 24,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6-sol": {
      "model_slug": "openai/gpt-6-sol",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "sakana/fugu-ultra-v2": {
      "model_slug": "sakana/fugu-ultra-v2",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "xiaomi/mimo-v2.6-pro-ultraspeed": {
      "model_slug": "xiaomi/mimo-v2.6-pro-ultraspeed",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inception/mercury-2.5": {
      "model_slug": "inception/mercury-2.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4.1-flash": {
      "model_slug": "deepseek/deepseek-v4.1-flash",
      "cosmos_test_count": 24,
      "pass_rate_pct": 87.5,
      "twelve_cs_score": 87.6,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=deepseek/deepseek-v4.1-flash": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "openai/gpt-6-astra-pro": {
      "model_slug": "openai/gpt-6-astra-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "xiaomi/mimo-v2.6-pro": {
      "model_slug": "xiaomi/mimo-v2.6-pro",
      "cosmos_test_count": 58,
      "pass_rate_pct": 10.3,
      "twelve_cs_score": 14.3,
      "passes": 6,
      "fails": 52,
      "rate_limits_429": 24,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'xiaomi/mimo-v2.6-pro'; want ['bytedance-seed/seed-2.|PIN-NOTE": 24,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"xiaomi/mimo-v2.6-pro is temporarily rate-limited upstream. Please retry sh": 8,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"xiaomi/mimo-v2": 8,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-6-astra": {
      "model_slug": "openai/gpt-6-astra",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "prism-ml/ternary-bonsai-2-27b": {
      "model_slug": "prism-ml/ternary-bonsai-2-27b",
      "cosmos_test_count": 34,
      "pass_rate_pct": 64.7,
      "twelve_cs_score": 66.0,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 9,
        "fizz-16lines-ok lines=16": 8,
        "sku=prism-ml/ternary-bonsai-2-27b": 4,
        "fizz-lines=16 first='NONE'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/gemini-3.8-flash": {
      "model_slug": "google/gemini-3.8-flash",
      "cosmos_test_count": 140,
      "pass_rate_pct": 12.1,
      "twelve_cs_score": 16.0,
      "passes": 17,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 6,
        "sku=google/gemini-3.8-flash": 4,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "z-ai/glm-5.3-flashx": {
      "model_slug": "z-ai/glm-5.3-flashx",
      "cosmos_test_count": 46,
      "pass_rate_pct": 30.4,
      "twelve_cs_score": 33.4,
      "passes": 14,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'z-ai/glm-5.3-flashx'; want ['bytedance-seed/seed-2.0|PIN-NOTE": 28,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "inference-net/schematron-v2-turbo": {
      "model_slug": "inference-net/schematron-v2-turbo",
      "cosmos_test_count": 27,
      "pass_rate_pct": 11.1,
      "twelve_cs_score": 15.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 24
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~openai/gpt-terra-latest": {
      "model_slug": "~openai/gpt-terra-latest",
      "cosmos_test_count": 18,
      "pass_rate_pct": 16.7,
      "twelve_cs_score": 20.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inclusionai/ling-3.0-flash-fin": {
      "model_slug": "inclusionai/ling-3.0-flash-fin",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~deepseek/deepseek-flash-latest": {
      "model_slug": "~deepseek/deepseek-flash-latest",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inclusionai/ling-3.0-flash-vl": {
      "model_slug": "inclusionai/ling-3.0-flash-vl",
      "cosmos_test_count": 27,
      "pass_rate_pct": 88.9,
      "twelve_cs_score": 89.0,
      "passes": 24,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=inclusionai/ling-3.0-flash-vl": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "meta/muse-spark-1.2-contributor": {
      "model_slug": "meta/muse-spark-1.2-contributor",
      "cosmos_test_count": 34,
      "pass_rate_pct": 73.5,
      "twelve_cs_score": 74.3,
      "passes": 25,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "EMPTY": 6,
        "mouth-form": 4,
        "sku=meta/muse-spark-1.2-contributor": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "~openai/gpt-sol-latest": {
      "model_slug": "~openai/gpt-sol-latest",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta/muse-spark-1.3": {
      "model_slug": "meta/muse-spark-1.3",
      "cosmos_test_count": 30,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 30,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "1": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "sakana/fugu-max": {
      "model_slug": "sakana/fugu-max",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nex-agi/nex-n2.5-mini": {
      "model_slug": "nex-agi/nex-n2.5-mini",
      "cosmos_test_count": 29,
      "pass_rate_pct": 48.3,
      "twelve_cs_score": 50.4,
      "passes": 14,
      "fails": 12,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http-detail:not a pinned OpenRouter id 'nex-agi/nex-n2.5-mini'; want ['bytedance-seed/seed-2": 8,
        "UNPINNED-NEEDS-PIN: OR /models lists nex-agi/nex-n2.5-mini LIVE (free twin gone); rail PINNED lacks it so require_duds/model_refused REFUSEs before HTTP; propose CCr pin, then smoke ping+fizz": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "~z-ai/glm-latest": {
      "model_slug": "~z-ai/glm-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.8-max-0902": {
      "model_slug": "qwen/qwen3.8-max-0902",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=qwen/qwen3.8-max-0902": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.7-flash": {
      "model_slug": "google/gemini-3.7-flash",
      "cosmos_test_count": 30,
      "pass_rate_pct": 73.3,
      "twelve_cs_score": 74.1,
      "passes": 22,
      "fails": 8,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\n  \\\"error\\\": {\\n    \\\"code\\\": 400,\\n    \\": 4,
        "print(\"NONE\")": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\n  \\\"error\\\": {\\n    \\": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-fable-5.1": {
      "model_slug": "anthropic/claude-fable-5.1",
      "cosmos_test_count": 15,
      "pass_rate_pct": 40.0,
      "twelve_cs_score": 42.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 6,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nex-agi/nex-n2.5-pro": {
      "model_slug": "nex-agi/nex-n2.5-pro",
      "cosmos_test_count": 26,
      "pass_rate_pct": 53.8,
      "twelve_cs_score": 55.6,
      "passes": 14,
      "fails": 12,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http-detail:not a pinned OpenRouter id 'nex-agi/nex-n2.5-pro'; want ['bytedance-seed/seed-2.": 8,
        "UNPINNED-NEEDS-PIN: OR /models lists nex-agi/nex-n2.5-pro LIVE (free twin gone); rail PINNED lacks it so require_duds/model_refused REFUSEs before HTTP; propose CCr pin, then smoke ping+fizz": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "tencent/hy4-preview": {
      "model_slug": "tencent/hy4-preview",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3,
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta/muse-spark-1.3-contributor": {
      "model_slug": "meta/muse-spark-1.3-contributor",
      "cosmos_test_count": 148,
      "pass_rate_pct": 20.9,
      "twelve_cs_score": 24.4,
      "passes": 31,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "NO_CONTEXT": 6,
        "sku=meta/muse-spark-1.3-contributor": 4,
        "lines=1 first='Resolving instruction hierarchy and outp'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "~z-ai/glm-flash-latest": {
      "model_slug": "~z-ai/glm-flash-latest",
      "cosmos_test_count": 19,
      "pass_rate_pct": 15.8,
      "twelve_cs_score": 19.5,
      "passes": 3,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 12,
        "ROTATOR-REFUSED|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5.3-flash": {
      "model_slug": "z-ai/glm-5.3-flash",
      "cosmos_test_count": 867,
      "pass_rate_pct": 2.8,
      "twelve_cs_score": 7.2,
      "passes": 24,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=z-ai/glm-5.3-flash": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "tencent/hy-mt2-30b-a3b": {
      "model_slug": "tencent/hy-mt2-30b-a3b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4-pro-0813": {
      "model_slug": "deepseek/deepseek-v4-pro-0813",
      "cosmos_test_count": 14,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 14,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "sku=deepseek/deepseek-v4-pro-0813": 4,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "upstage/solar-pro4": {
      "model_slug": "upstage/solar-pro4",
      "cosmos_test_count": 30,
      "pass_rate_pct": 70.0,
      "twelve_cs_score": 71.0,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=upstage/solar-pro4": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "ibm-granite/granite-4.2-8b": {
      "model_slug": "ibm-granite/granite-4.2-8b",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4-flash-vision-exp": {
      "model_slug": "deepseek/deepseek-v4-flash-vision-exp",
      "cosmos_test_count": 49,
      "pass_rate_pct": 28.6,
      "twelve_cs_score": 31.7,
      "passes": 14,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'deepseek/deepseek-v4-flash-vision-exp'; want ['byted|PIN-NOTE": 28,
        "revisit-later|PIN-NOTE": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "tencent/hy-mt2-7b": {
      "model_slug": "tencent/hy-mt2-7b",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4-flash-0731": {
      "model_slug": "deepseek/deepseek-v4-flash-0731",
      "cosmos_test_count": 138,
      "pass_rate_pct": 15.2,
      "twelve_cs_score": 18.9,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 12,
        "fizz-16lines-ok lines=16": 8,
        "sku=deepseek/deepseek-v4-flash-0731": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "inclusionai/ling-3.0-flash": {
      "model_slug": "inclusionai/ling-3.0-flash",
      "cosmos_test_count": 30,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 30,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "sku=inclusionai/ling-3.0-flash": 8,
        "fizz-16lines-ok lines=16": 8,
        "fizz-lines=33 first='1.  **Analyze the Request:**'": 4,
        "fizz-lines=27 first='1.  **Analyze the Request:**'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.8-flash": {
      "model_slug": "qwen/qwen3.8-flash",
      "cosmos_test_count": 132,
      "pass_rate_pct": 15.9,
      "twelve_cs_score": 19.6,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 12,
        "fizz-16lines-ok lines=16": 8,
        "sku=qwen/qwen3.8-flash": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "tencent/hy-mt2-1.8b": {
      "model_slug": "tencent/hy-mt2-1.8b",
      "cosmos_test_count": 19,
      "pass_rate_pct": 15.8,
      "twelve_cs_score": 19.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs",
        "_hero_run_20261008"
      ]
    },
    "qwen/qwen3.8-27b": {
      "model_slug": "qwen/qwen3.8-27b",
      "cosmos_test_count": 27,
      "pass_rate_pct": 77.8,
      "twelve_cs_score": 78.4,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=qwen/qwen3.8-27b": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "z-ai/glm-5.3": {
      "model_slug": "z-ai/glm-5.3",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=z-ai/glm-5.3": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-2-1-turbo": {
      "model_slug": "bytedance-seed/seed-2-1-turbo",
      "cosmos_test_count": 48,
      "pass_rate_pct": 22.9,
      "twelve_cs_score": 26.3,
      "passes": 11,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'bytedance-seed/seed-2-1-turbo'; want ['bytedance-see|PIN-NOTE": 24,
        "EMPTY": 9,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "meituan/longcat-2.0": {
      "model_slug": "meituan/longcat-2.0",
      "cosmos_test_count": 45,
      "pass_rate_pct": 24.4,
      "twelve_cs_score": 27.7,
      "passes": 11,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'meituan/longcat-2.0'; want ['bytedance-seed/seed-2.0|PIN-NOTE": 24,
        "revisit-later|PIN-NOTE": 4,
        "HOST_FAIL": 3,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "meta/muse-spark-1.1": {
      "model_slug": "meta/muse-spark-1.1",
      "cosmos_test_count": 26,
      "pass_rate_pct": 69.2,
      "twelve_cs_score": 70.2,
      "passes": 18,
      "fails": 8,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "<empty>": 12,
        "not a pinned OpenRouter id 'meta/muse-spark-1.1'; want ['bytedance-seed/seed-2.0": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-2.0-code": {
      "model_slug": "bytedance-seed/seed-2.0-code",
      "cosmos_test_count": 49,
      "pass_rate_pct": 22.4,
      "twelve_cs_score": 25.8,
      "passes": 11,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'bytedance-seed/seed-2.0-code'; want ['bytedance-seed|PIN-NOTE": 28,
        "EMPTY": 6,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "x-ai/grok-4.6": {
      "model_slug": "x-ai/grok-4.6",
      "cosmos_test_count": 26,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "mouth-form": 12,
        "sku=x-ai/grok-4.6": 4,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "sakana/sakana-namazu": {
      "model_slug": "sakana/sakana-namazu",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.6-terra-pro": {
      "model_slug": "openai/gpt-5.6-terra-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.8-2.4t-a95b": {
      "model_slug": "qwen/qwen3.8-2.4t-a95b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nvidia/nemotron-3.5-lightning": {
      "model_slug": "nvidia/nemotron-3.5-lightning",
      "cosmos_test_count": 38,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 38,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-lines=15 first='Fizz'": 8,
        "<empty>": 8,
        "1": 8,
        "sku=nvidia/nemotron-3.5-lightning": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "~deepseek/deepseek-v4-flash-latest": {
      "model_slug": "~deepseek/deepseek-v4-flash-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "x-ai/grok-4.5": {
      "model_slug": "x-ai/grok-4.5",
      "cosmos_test_count": 14,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 14,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-5": {
      "model_slug": "anthropic/claude-opus-5",
      "cosmos_test_count": 18,
      "pass_rate_pct": 77.8,
      "twelve_cs_score": 78.4,
      "passes": 14,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4,
        "reply-line-unverified": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "meta/muse-glimmer-30b": {
      "model_slug": "meta/muse-glimmer-30b",
      "cosmos_test_count": 62,
      "pass_rate_pct": 48.4,
      "twelve_cs_score": 50.5,
      "passes": 30,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'meta/muse-glimmer-30b'; want ['bytedance-seed/seed-2|PIN-NOTE": 28,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4,
        "EMPTY-ON-LONG finish=length": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "meta/muse-spark-1.2": {
      "model_slug": "meta/muse-spark-1.2",
      "cosmos_test_count": 25,
      "pass_rate_pct": 88.0,
      "twelve_cs_score": 88.1,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "mouth-form": 4,
        "sku=meta/muse-spark-1.2": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.5-flash-lite": {
      "model_slug": "google/gemini-3.5-flash-lite",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "tencent/hy3": {
      "model_slug": "tencent/hy3",
      "cosmos_test_count": 14,
      "pass_rate_pct": 71.4,
      "twelve_cs_score": 72.3,
      "passes": 10,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "PIN-HOLE|PIN-NOTE": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "thinkingmachines/inkling-small": {
      "model_slug": "thinkingmachines/inkling-small",
      "cosmos_test_count": 46,
      "pass_rate_pct": 30.4,
      "twelve_cs_score": 33.4,
      "passes": 14,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'thinkingmachines/inkling-small'; want ['bytedance-se|PIN-NOTE": 28,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.7-flash": {
      "model_slug": "qwen/qwen3.7-flash",
      "cosmos_test_count": 45,
      "pass_rate_pct": 57.8,
      "twelve_cs_score": 59.4,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 9,
        "sku=qwen/qwen3.7-flash": 4,
        "fizz-lines=89 first='The user wants a specific output format:'": 4,
        "fizz-lines=15 first='NONE'": 4
      },
      "source_clusters": [
        "womb_scars",
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "roster"
      ]
    },
    "moonshotai/kimi-k3": {
      "model_slug": "moonshotai/kimi-k3",
      "cosmos_test_count": 24,
      "pass_rate_pct": 75.0,
      "twelve_cs_score": 75.8,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 6,
        "The intent is a simple FizzBuzz script: print \"NONE\" as the first line": 4,
        "reply-line-unverified": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.6-luna": {
      "model_slug": "openai/gpt-5.6-luna",
      "cosmos_test_count": 22,
      "pass_rate_pct": 81.8,
      "twelve_cs_score": 82.2,
      "passes": 18,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-5.6-luna": 4,
        "lines=1 first=\"http=200 error='Flex processing is tempo\"": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "poolside/laguna-s-2.1": {
      "model_slug": "poolside/laguna-s-2.1",
      "cosmos_test_count": 28,
      "pass_rate_pct": 89.3,
      "twelve_cs_score": 89.3,
      "passes": 25,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=poolside/laguna-s-2.1": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "openai/gpt-5.6-sol": {
      "model_slug": "openai/gpt-5.6-sol",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-5.6-sol": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "thinkingmachines/inkling": {
      "model_slug": "thinkingmachines/inkling",
      "cosmos_test_count": 64,
      "pass_rate_pct": 90.6,
      "twelve_cs_score": 90.6,
      "passes": 58,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "mouth-form": 20,
        "fizz-lines=1 first='NONE12Fizz4BuzzFizz78FizzBuzz11Fizz1314F'": 12,
        "sku=thinkingmachines/inkling": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "kwaipilot/kat-coder-pro-v2.5": {
      "model_slug": "kwaipilot/kat-coder-pro-v2.5",
      "cosmos_test_count": 21,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 21
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.6-terra": {
      "model_slug": "openai/gpt-5.6-terra",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-5.6-terra": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-3.0": {
      "model_slug": "aion-labs/aion-3.0",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-flash-lite-image": {
      "model_slug": "google/gemini-3.1-flash-lite-image",
      "cosmos_test_count": 15,
      "pass_rate_pct": 40.0,
      "twelve_cs_score": 42.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~x-ai/grok-latest": {
      "model_slug": "~x-ai/grok-latest",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5.2": {
      "model_slug": "z-ai/glm-5.2",
      "cosmos_test_count": 22,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=z-ai/glm-5.2": 4,
        "lines=1 first='Written `fizzbuzz.py` \u2014 running it produ'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "sakana/fugu-ultra": {
      "model_slug": "sakana/fugu-ultra",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-fable-5": {
      "model_slug": "anthropic/claude-fable-5",
      "cosmos_test_count": 18,
      "pass_rate_pct": 16.7,
      "twelve_cs_score": 20.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openrouter/fusion": {
      "model_slug": "openrouter/fusion",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nvidia/nemotron-3.5-content-safety": {
      "model_slug": "nvidia/nemotron-3.5-content-safety",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 18
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "poolside/laguna-xs-2.1": {
      "model_slug": "poolside/laguna-xs-2.1",
      "cosmos_test_count": 21,
      "pass_rate_pct": 85.7,
      "twelve_cs_score": 85.9,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=poolside/laguna-xs-2.1": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-flash-image": {
      "model_slug": "google/gemini-3.1-flash-image",
      "cosmos_test_count": 15,
      "pass_rate_pct": 40.0,
      "twelve_cs_score": 42.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "stepfun/step-3.7-flash": {
      "model_slug": "stepfun/step-3.7-flash",
      "cosmos_test_count": 49,
      "pass_rate_pct": 22.4,
      "twelve_cs_score": 25.8,
      "passes": 11,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'stepfun/step-3.7-flash'; want ['bytedance-seed/seed-|PIN-NOTE": 28,
        "EMPTY": 6,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "moonshotai/kimi-k2.7-code": {
      "model_slug": "moonshotai/kimi-k2.7-code",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nvidia/nemotron-3-ultra-550b-a55b": {
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b",
      "cosmos_test_count": 25,
      "pass_rate_pct": 88.0,
      "twelve_cs_score": 88.1,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "sku=nvidia/nemotron-3-ultra-550b-a55b": 4,
        "fizz-16lines-ok lines=16": 4,
        "lines=15 first='1'": 4,
        "lines=1 first='Done. The `fizzbuzz.py` file outputs the'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/gemini-3.5-flash": {
      "model_slug": "google/gemini-3.5-flash",
      "cosmos_test_count": 30,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 30,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-4.8": {
      "model_slug": "anthropic/claude-opus-4.8",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "perceptron/perceptron-mk1": {
      "model_slug": "perceptron/perceptron-mk1",
      "cosmos_test_count": 42,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 14,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'perceptron/perceptron-mk1'; want ['bytedance-seed/se|PIN-NOTE": 24,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-medium-3-5": {
      "model_slug": "mistralai/mistral-medium-3-5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "x-ai/grok-4.3": {
      "model_slug": "x-ai/grok-4.3",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~moonshotai/kimi-latest": {
      "model_slug": "~moonshotai/kimi-latest",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9,
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m3": {
      "model_slug": "minimax/minimax-m3",
      "cosmos_test_count": 45,
      "pass_rate_pct": 24.4,
      "twelve_cs_score": 27.7,
      "passes": 11,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'minimax/minimax-m3'; want ['bytedance-seed/seed-2.0-|PIN-NOTE": 24,
        "EMPTY": 6,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "~google/gemini-pro-latest": {
      "model_slug": "~google/gemini-pro-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.6-flash": {
      "model_slug": "qwen/qwen3.6-flash",
      "cosmos_test_count": 48,
      "pass_rate_pct": 22.9,
      "twelve_cs_score": 26.3,
      "passes": 11,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'qwen/qwen3.6-flash'; want ['bytedance-seed/seed-2.0-|PIN-NOTE": 24,
        "EMPTY": 9,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "x-ai/grok-build-0.1": {
      "model_slug": "x-ai/grok-build-0.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-chat-latest": {
      "model_slug": "openai/gpt-chat-latest",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~openai/gpt-mini-latest": {
      "model_slug": "~openai/gpt-mini-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~anthropic/claude-sonnet-latest": {
      "model_slug": "~anthropic/claude-sonnet-latest",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-plus-20260420": {
      "model_slug": "qwen/qwen3.5-plus-20260420",
      "cosmos_test_count": 69,
      "pass_rate_pct": 50.7,
      "twelve_cs_score": 52.7,
      "passes": 35,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'qwen/qwen3.5-plus-20260420'; want ['bytedance-seed/s|PIN-NOTE": 24,
        "<empty>": 8,
        "EMPTY": 6,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.6-max-preview": {
      "model_slug": "qwen/qwen3.6-max-preview",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.5-pro": {
      "model_slug": "openai/gpt-5.5-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4-pro": {
      "model_slug": "deepseek/deepseek-v4-pro",
      "cosmos_test_count": 26,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "empty": 4,
        "```python": 4,
        "print(\"NONE\")": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.6-27b": {
      "model_slug": "qwen/qwen3.6-27b",
      "cosmos_test_count": 46,
      "pass_rate_pct": 39.1,
      "twelve_cs_score": 41.6,
      "passes": 18,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'qwen/qwen3.6-27b'; want ['bytedance-seed/seed-2.0-mi|PIN-NOTE": 24,
        "<empty>": 8,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "tencent/hy3-preview": {
      "model_slug": "tencent/hy3-preview",
      "cosmos_test_count": 998,
      "pass_rate_pct": 4.3,
      "twelve_cs_score": 8.6,
      "passes": 43,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 6,
        "sku=tencent/hy3-preview": 4,
        "fizz-16lines-ok lines=16": 4,
        "lines=56 first='The user wants me to reply with exactly '": 4
      },
      "source_clusters": [
        "womb_scars",
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "roster"
      ]
    },
    "~anthropic/claude-opus-latest": {
      "model_slug": "~anthropic/claude-opus-latest",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5.1": {
      "model_slug": "z-ai/glm-5.1",
      "cosmos_test_count": 14,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 14,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "xiaomi/mimo-v2.5": {
      "model_slug": "xiaomi/mimo-v2.5",
      "cosmos_test_count": 36,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 24,
      "fails": 3,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=xiaomi/mimo-v2.5": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "moonshotai/kimi-k2.6": {
      "model_slug": "moonshotai/kimi-k2.6",
      "cosmos_test_count": 55,
      "pass_rate_pct": 27.3,
      "twelve_cs_score": 30.4,
      "passes": 15,
      "fails": 28,
      "rate_limits_429": 12,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'moonshotai/kimi-k2.6'; want ['bytedance-seed/seed-2.|PIN-NOTE": 24,
        "UPSTREAM_POOL": 12,
        " NONE": 12,
        "revisit-later|PIN-NOTE": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5v-turbo": {
      "model_slug": "z-ai/glm-5v-turbo",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-4-31b-it": {
      "model_slug": "google/gemma-4-31b-it",
      "cosmos_test_count": 58,
      "pass_rate_pct": 84.5,
      "twelve_cs_score": 84.8,
      "passes": 49,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-lines=15 first='1'": 8,
        "<empty>": 8,
        "1": 8,
        "UPSTREAM_POOL": 6
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "x-ai/grok-4.20-multi-agent": {
      "model_slug": "x-ai/grok-4.20-multi-agent",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.4-nano": {
      "model_slug": "openai/gpt-5.4-nano",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v4-flash": {
      "model_slug": "deepseek/deepseek-v4-flash",
      "cosmos_test_count": 30,
      "pass_rate_pct": 70.0,
      "twelve_cs_score": 71.0,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "HOST_FAIL": 6,
        "sku=deepseek/deepseek-v4-flash": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3-super-120b-a12b": {
      "model_slug": "nvidia/nemotron-3-super-120b-a12b",
      "cosmos_test_count": 7,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 7,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs",
        "_hero_run_20261008"
      ]
    },
    "openai/gpt-5.4": {
      "model_slug": "openai/gpt-5.4",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-flash-image-preview": {
      "model_slug": "google/gemini-3.1-flash-image-preview",
      "cosmos_test_count": 12,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-flash-02-23": {
      "model_slug": "qwen/qwen3.5-flash-02-23",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-pro-preview": {
      "model_slug": "google/gemini-3.1-pro-preview",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m2.5": {
      "model_slug": "minimax/minimax-m2.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-coder-next": {
      "model_slug": "qwen/qwen3-coder-next",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.4-image-2": {
      "model_slug": "openai/gpt-5.4-image-2",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-4.7": {
      "model_slug": "anthropic/claude-opus-4.7",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m2-her": {
      "model_slug": "minimax/minimax-m2-her",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.6-plus": {
      "model_slug": "qwen/qwen3.6-plus",
      "cosmos_test_count": 57,
      "pass_rate_pct": 40.4,
      "twelve_cs_score": 42.9,
      "passes": 23,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'qwen/qwen3.6-plus'; want ['bytedance-seed/seed-2.0-m|PIN-NOTE": 24,
        "EMPTY": 6,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "x-ai/grok-4.20": {
      "model_slug": "x-ai/grok-4.20",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.4-mini": {
      "model_slug": "openai/gpt-5.4-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.7-flash": {
      "model_slug": "z-ai/glm-4.7-flash",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-2.0-lite": {
      "model_slug": "bytedance-seed/seed-2.0-lite",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "inception/mercury-2": {
      "model_slug": "inception/mercury-2",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-35b-a3b": {
      "model_slug": "qwen/qwen3.5-35b-a3b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m2.1": {
      "model_slug": "minimax/minimax-m2.1",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.2-chat": {
      "model_slug": "openai/gpt-5.2-chat",
      "cosmos_test_count": 12,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "relace/relace-search": {
      "model_slug": "relace/relace-search",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-pro-preview-customtools": {
      "model_slug": "google/gemini-3.1-pro-preview-customtools",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "amazon/nova-2-lite-v1": {
      "model_slug": "amazon/nova-2-lite-v1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-large-2512": {
      "model_slug": "mistralai/mistral-large-2512",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-4.6": {
      "model_slug": "anthropic/claude-sonnet-4.6",
      "cosmos_test_count": 14,
      "pass_rate_pct": 71.4,
      "twelve_cs_score": 72.3,
      "passes": 10,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.1": {
      "model_slug": "openai/gpt-5.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "amazon/nova-premier-v1": {
      "model_slug": "amazon/nova-premier-v1",
      "cosmos_test_count": 12,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m2": {
      "model_slug": "minimax/minimax-m2",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5": {
      "model_slug": "z-ai/glm-5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-haiku-4.5": {
      "model_slug": "anthropic/claude-haiku-4.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-2.5-flash-image": {
      "model_slug": "google/gemini-2.5-flash-image",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.6": {
      "model_slug": "z-ai/glm-4.6",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "relace/relace-apply-3": {
      "model_slug": "relace/relace-apply-3",
      "cosmos_test_count": 18,
      "pass_rate_pct": 16.7,
      "twelve_cs_score": 20.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-coder-plus": {
      "model_slug": "qwen/qwen3-coder-plus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-next-80b-a3b-instruct": {
      "model_slug": "qwen/qwen3-next-80b-a3b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "stepfun/step-3.5-flash": {
      "model_slug": "stepfun/step-3.5-flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nousresearch/hermes-4-405b": {
      "model_slug": "nousresearch/hermes-4-405b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "writer/palmyra-x5": {
      "model_slug": "writer/palmyra-x5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.2-codex": {
      "model_slug": "openai/gpt-5.2-codex",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5": {
      "model_slug": "openai/gpt-5",
      "cosmos_test_count": 14,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 14,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-oss-20b": {
      "model_slug": "openai/gpt-oss-20b",
      "cosmos_test_count": 28,
      "pass_rate_pct": 89.3,
      "twelve_cs_score": 89.3,
      "passes": 25,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-oss-20b": 4,
        "fizz-lines=16 first='NONE'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "qwen/qwen3-30b-a3b-instruct-2507": {
      "model_slug": "qwen/qwen3-30b-a3b-instruct-2507",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-coder": {
      "model_slug": "qwen/qwen3-coder",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "moonshotai/kimi-k2": {
      "model_slug": "moonshotai/kimi-k2",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK kimi-k2": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "morph/morph-v3-fast": {
      "model_slug": "morph/morph-v3-fast",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-2.5-flash": {
      "model_slug": "google/gemini-2.5-flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.7": {
      "model_slug": "z-ai/glm-4.7",
      "cosmos_test_count": 30,
      "pass_rate_pct": 90.0,
      "twelve_cs_score": 90.0,
      "passes": 27,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 8,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.2-pro": {
      "model_slug": "openai/gpt-5.2-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.6v": {
      "model_slug": "z-ai/glm-4.6v",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/ministral-14b-2512": {
      "model_slug": "mistralai/ministral-14b-2512",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v3.2": {
      "model_slug": "deepseek/deepseek-v3.2",
      "cosmos_test_count": 34,
      "pass_rate_pct": 88.2,
      "twelve_cs_score": 88.3,
      "passes": 30,
      "fails": 4,
      "rate_limits_429": 4,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"deepseek/deepseek-v3.2 is temporarily rate-l": 4,
        "OCEAN_3 deepseek-v3.2": 4,
        "OVERRIDE_INVALID deepseek-v3.2": 4,
        "package design.factory.method;": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.1-codex": {
      "model_slug": "openai/gpt-5.1-codex",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "perplexity/sonar-pro-search": {
      "model_slug": "perplexity/sonar-pro-search",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-32b-instruct": {
      "model_slug": "qwen/qwen3-vl-32b-instruct",
      "cosmos_test_count": 46,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 46,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-lines=15 first='1'": 8,
        "<empty>": 8,
        "sku=qwen/qwen3-vl-32b-instruct": 4,
        "fizz-lines=15 first='NONE'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-8b-thinking": {
      "model_slug": "qwen/qwen3-vl-8b-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-30b-a3b-thinking": {
      "model_slug": "qwen/qwen3-vl-30b-a3b-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-4.5": {
      "model_slug": "anthropic/claude-sonnet-4.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-r1-0528": {
      "model_slug": "deepseek/deepseek-r1-0528",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-30b-a3b": {
      "model_slug": "qwen/qwen3-30b-a3b",
      "cosmos_test_count": 34,
      "pass_rate_pct": 91.2,
      "twelve_cs_score": 91.1,
      "passes": 31,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "sku=qwen/qwen3-30b-a3b": 4,
        "fizz-16lines-ok lines=16": 4,
        "lines=20 first='The Python script has been written to pr'": 4,
        "lines=15 first='1'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-235b-a22b": {
      "model_slug": "qwen/qwen3-235b-a22b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-235b-a22b-thinking": {
      "model_slug": "qwen/qwen3-vl-235b-a22b-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4.1": {
      "model_slug": "openai/gpt-4.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-4-scout": {
      "model_slug": "meta-llama/llama-4-scout",
      "cosmos_test_count": 55,
      "pass_rate_pct": 5.5,
      "twelve_cs_score": 9.7,
      "passes": 3,
      "fails": 43,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "404-dead-slug": 28,
        "SLUG_DEAD": 9,
        "<empty>": 4,
        "http=404 error='No endpoints found for meta-llama/llama-4-scout:floor. Every can": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "deepseek/deepseek-v3.1-terminus": {
      "model_slug": "deepseek/deepseek-v3.1-terminus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen-plus-2025-07-28": {
      "model_slug": "qwen/qwen-plus-2025-07-28",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-3-4b-it": {
      "model_slug": "google/gemma-3-4b-it",
      "cosmos_test_count": 35,
      "pass_rate_pct": 62.9,
      "twelve_cs_score": 64.3,
      "passes": 22,
      "fails": 13,
      "rate_limits_429": 8,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK gemma-3-4b-it": 20,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"google/gemma-3-4b-it is temporarily rate-limited upstream. Please retry sh": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"google/gemma-3-4b-it is ": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-chat-v3.1": {
      "model_slug": "deepseek/deepseek-chat-v3.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-3-27b-it": {
      "model_slug": "google/gemma-3-27b-it",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```python": 4,
        "1": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5-mini": {
      "model_slug": "openai/gpt-5-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-4.1": {
      "model_slug": "anthropic/claude-opus-4.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.5": {
      "model_slug": "z-ai/glm-4.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "bytedance/ui-tars-1.5-7b": {
      "model_slug": "bytedance/ui-tars-1.5-7b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "cognitivecomputations/dolphin-mistral-24b-venice-edition": {
      "model_slug": "cognitivecomputations/dolphin-mistral-24b-venice-edition",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "baidu/ernie-4.5-vl-424b-a47b": {
      "model_slug": "baidu/ernie-4.5-vl-424b-a47b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "perplexity/sonar-deep-research": {
      "model_slug": "perplexity/sonar-deep-research",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen2.5-vl-72b-instruct": {
      "model_slug": "qwen/qwen2.5-vl-72b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "perplexity/sonar": {
      "model_slug": "perplexity/sonar",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-chat": {
      "model_slug": "deepseek/deepseek-chat",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-3.3-70b-instruct": {
      "model_slug": "meta-llama/llama-3.3-70b-instruct",
      "cosmos_test_count": 21,
      "pass_rate_pct": 71.4,
      "twelve_cs_score": 72.3,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "SLUG_DEAD": 6,
        "sku=meta-llama/llama-3.3-70b-instruct": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-2.5-pro": {
      "model_slug": "google/gemini-2.5-pro",
      "cosmos_test_count": 18,
      "pass_rate_pct": 83.3,
      "twelve_cs_score": 83.6,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o-2024-11-20": {
      "model_slug": "openai/gpt-4o-2024-11-20",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-4": {
      "model_slug": "anthropic/claude-sonnet-4",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthracite-org/magnum-v4-72b": {
      "model_slug": "anthracite-org/magnum-v4-72b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen-2.5-72b-instruct": {
      "model_slug": "qwen/qwen-2.5-72b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nousresearch/hermes-3-llama-3.1-70b": {
      "model_slug": "nousresearch/hermes-3-llama-3.1-70b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-3.1-70b-instruct": {
      "model_slug": "meta-llama/llama-3.1-70b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o-mini-2024-07-18": {
      "model_slug": "openai/gpt-4o-mini-2024-07-18",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mixtral-8x22b-instruct": {
      "model_slug": "mistralai/mixtral-8x22b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-3.5-turbo-0613": {
      "model_slug": "openai/gpt-3.5-turbo-0613",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "undi95/remm-slerp-l2-13b": {
      "model_slug": "undi95/remm-slerp-l2-13b",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-8b": {
      "model_slug": "qwen/qwen3-8b",
      "cosmos_test_count": 40,
      "pass_rate_pct": 85.0,
      "twelve_cs_score": 85.2,
      "passes": 34,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 6,
        "sku=qwen/qwen3-8b": 4,
        "fizz-16lines-ok lines=16": 4,
        "lines=22 first='The correct FizzBuzz sequence has been g'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/o4-mini-high": {
      "model_slug": "openai/o4-mini-high",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4.1-mini": {
      "model_slug": "openai/gpt-4.1-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-chat-v3-0324": {
      "model_slug": "deepseek/deepseek-chat-v3-0324",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-3-12b-it": {
      "model_slug": "google/gemma-3-12b-it",
      "cosmos_test_count": 19,
      "pass_rate_pct": 78.9,
      "twelve_cs_score": 79.5,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```python": 4,
        "1": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "thedrummer/skyfall-36b-v2": {
      "model_slug": "thedrummer/skyfall-36b-v2",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-saba": {
      "model_slug": "mistralai/mistral-saba",
      "cosmos_test_count": 26,
      "pass_rate_pct": 11.5,
      "twelve_cs_score": 15.4,
      "passes": 3,
      "fails": 20,
      "rate_limits_429": 23,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"mistralai/mistral-saba is temporarily rate-limited upstream. Please retry ": 8,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"mistralai/mist": 8,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"mistralai/mistral-saba i": 4,
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen-plus": {
      "model_slug": "qwen/qwen-plus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-r1": {
      "model_slug": "deepseek/deepseek-r1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "sao10k/l3.3-euryale-70b": {
      "model_slug": "sao10k/l3.3-euryale-70b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "amazon/nova-lite-v1": {
      "model_slug": "amazon/nova-lite-v1",
      "cosmos_test_count": 22,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```python": 4,
        " ": 4,
        "<thinking>The user has requested a specific output format without providing a ta": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-large-2407": {
      "model_slug": "mistralai/mistral-large-2407",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen-2.5-7b-instruct": {
      "model_slug": "qwen/qwen-2.5-7b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "cohere/command-r-08-2024": {
      "model_slug": "cohere/command-r-08-2024",
      "cosmos_test_count": 22,
      "pass_rate_pct": 86.4,
      "twelve_cs_score": 86.6,
      "passes": 19,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```": 4,
        "HERO_OK command-r-08-2024": 4,
        "1": 4,
        "narrates-then-complies": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "nousresearch/hermes-3-llama-3.1-405b": {
      "model_slug": "nousresearch/hermes-3-llama-3.1-405b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-3.1-8b-instruct": {
      "model_slug": "meta-llama/llama-3.1-8b-instruct",
      "cosmos_test_count": 22,
      "pass_rate_pct": 86.4,
      "twelve_cs_score": 86.6,
      "passes": 19,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK llama-3.1-8b-instruct": 8,
        "<b>function=HERO_OK llama-3.1-8b-instruct</function>": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemma-2-27b-it": {
      "model_slug": "google/gemma-2-27b-it",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "microsoft/wizardlm-2-8x22b": {
      "model_slug": "microsoft/wizardlm-2-8x22b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-3.5-turbo-instruct": {
      "model_slug": "openai/gpt-3.5-turbo-instruct",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "gryphe/mythomax-l2-13b": {
      "model_slug": "gryphe/mythomax-l2-13b",
      "cosmos_test_count": 28,
      "pass_rate_pct": 53.6,
      "twelve_cs_score": 55.4,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 9,
        "HERO_OK mythomax-l2-13b": 8,
        " HERO_OK mythomax-l2-13b": 4,
        "\u001b[0m": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.6-flash": {
      "model_slug": "google/gemini-3.6-flash",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=google/gemini-3.6-flash": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openrouter/auto-beta": {
      "model_slug": "openrouter/auto-beta",
      "cosmos_test_count": 21,
      "pass_rate_pct": 14.3,
      "twelve_cs_score": 18.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SLUG_DEAD": 18
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.6-luna-pro": {
      "model_slug": "openai/gpt-5.6-luna-pro",
      "cosmos_test_count": 22,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 22,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-5.6-luna-pro": 4,
        "lines=16 first='print(\"NONE\")'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.6-sol-pro": {
      "model_slug": "openai/gpt-5.6-sol-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-3.0-mini": {
      "model_slug": "aion-labs/aion-3.0-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-sonnet-5": {
      "model_slug": "anthropic/claude-sonnet-5",
      "cosmos_test_count": 18,
      "pass_rate_pct": 77.8,
      "twelve_cs_score": 78.4,
      "passes": 14,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4,
        "reply-line-unverified": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "rekaai/reka-edge": {
      "model_slug": "rekaai/reka-edge",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-small-2603": {
      "model_slug": "mistralai/mistral-small-2603",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3-pro-image": {
      "model_slug": "google/gemini-3-pro-image",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~anthropic/claude-fable-latest": {
      "model_slug": "~anthropic/claude-fable-latest",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.7-plus": {
      "model_slug": "qwen/qwen3.7-plus",
      "cosmos_test_count": 42,
      "pass_rate_pct": 26.2,
      "twelve_cs_score": 29.4,
      "passes": 11,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'qwen/qwen3.7-plus'; want ['bytedance-seed/seed-2.0-m|PIN-NOTE": 24,
        "revisit-later|PIN-NOTE": 4,
        "<empty>": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.7-max": {
      "model_slug": "qwen/qwen3.7-max",
      "cosmos_test_count": 18,
      "pass_rate_pct": 83.3,
      "twelve_cs_score": 83.6,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-flash-lite": {
      "model_slug": "google/gemini-3.1-flash-lite",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~anthropic/claude-haiku-latest": {
      "model_slug": "~anthropic/claude-haiku-latest",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "~google/gemini-flash-latest": {
      "model_slug": "~google/gemini-flash-latest",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-9b": {
      "model_slug": "qwen/qwen3.5-9b",
      "cosmos_test_count": 18,
      "pass_rate_pct": 83.3,
      "twelve_cs_score": 83.6,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=qwen/qwen3.5-9b": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3.1-flash-lite-preview": {
      "model_slug": "google/gemini-3.1-flash-lite-preview",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.6-35b-a3b": {
      "model_slug": "qwen/qwen3.6-35b-a3b",
      "cosmos_test_count": 24,
      "pass_rate_pct": 62.5,
      "twelve_cs_score": 63.9,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=qwen/qwen3.6-35b-a3b": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "qwen/qwen3.5-27b": {
      "model_slug": "qwen/qwen3.5-27b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.5": {
      "model_slug": "openai/gpt-5.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.3-codex": {
      "model_slug": "openai/gpt-5.3-codex",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "xiaomi/mimo-v2.5-pro": {
      "model_slug": "xiaomi/mimo-v2.5-pro",
      "cosmos_test_count": 70,
      "pass_rate_pct": 38.6,
      "twelve_cs_score": 41.2,
      "passes": 27,
      "fails": 40,
      "rate_limits_429": 12,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'xiaomi/mimo-v2.5-pro'; want ['bytedance-seed/seed-2.|PIN-NOTE": 24,
        "HERO_OK mimo-v2.5-pro": 8,
        "revisit-later|PIN-NOTE": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"xiaomi/mimo-v2.5-pro is temporarily rate-lim": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-plus-02-15": {
      "model_slug": "qwen/qwen3.5-plus-02-15",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openrouter/pareto-code": {
      "model_slug": "openrouter/pareto-code",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-max-thinking": {
      "model_slug": "qwen/qwen3-max-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemma-4-26b-a4b-it": {
      "model_slug": "google/gemma-4-26b-a4b-it",
      "cosmos_test_count": 24,
      "pass_rate_pct": 87.5,
      "twelve_cs_score": 87.6,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=google/gemma-4-26b-a4b-it": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "arcee-ai/trinity-large-thinking": {
      "model_slug": "arcee-ai/trinity-large-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m2.7": {
      "model_slug": "minimax/minimax-m2.7",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-5-turbo": {
      "model_slug": "z-ai/glm-5-turbo",
      "cosmos_test_count": 14,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 14,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "moonshotai/kimi-k2.5": {
      "model_slug": "moonshotai/kimi-k2.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-audio": {
      "model_slug": "openai/gpt-audio",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-1.6-flash": {
      "model_slug": "bytedance-seed/seed-1.6-flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3-flash-preview": {
      "model_slug": "google/gemini-3-flash-preview",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.2": {
      "model_slug": "openai/gpt-5.2",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openrouter/bodybuilder": {
      "model_slug": "openrouter/bodybuilder",
      "cosmos_test_count": 18,
      "pass_rate_pct": 16.7,
      "twelve_cs_score": 20.4,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/ministral-8b-2512": {
      "model_slug": "mistralai/ministral-8b-2512",
      "cosmos_test_count": 35,
      "pass_rate_pct": 91.4,
      "twelve_cs_score": 91.3,
      "passes": 32,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "mouth-form": 4,
        "sku=mistralai/ministral-8b-2512": 4,
        "lines=9 first='NONE'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "anthropic/claude-opus-4.5": {
      "model_slug": "anthropic/claude-opus-4.5",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.1-codex-mini": {
      "model_slug": "openai/gpt-5.1-codex-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/voxtral-small-24b-2507": {
      "model_slug": "mistralai/voxtral-small-24b-2507",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "ibm-granite/granite-4.0-h-micro": {
      "model_slug": "ibm-granite/granite-4.0-h-micro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-8b-instruct": {
      "model_slug": "qwen/qwen3-vl-8b-instruct",
      "cosmos_test_count": 46,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 46,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-lines=15 first='1'": 12,
        "<empty>": 8,
        "1": 8,
        "sku=qwen/qwen3-vl-8b-instruct": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-vl-30b-a3b-instruct": {
      "model_slug": "qwen/qwen3-vl-30b-a3b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "deepseek/deepseek-v3.2-exp": {
      "model_slug": "deepseek/deepseek-v3.2-exp",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.4-pro": {
      "model_slug": "openai/gpt-5.4-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-2.0-mini": {
      "model_slug": "bytedance-seed/seed-2.0-mini",
      "cosmos_test_count": 64,
      "pass_rate_pct": 95.3,
      "twelve_cs_score": 95.0,
      "passes": 61,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "mouth-form": 28,
        "<empty>": 8,
        "{\"l1_role\": false, \"l2_model\": true, \"l3_harness\": true, \"l4_wrapper\": true, \"l5_skills\": true, \"l6_tools\": true, \"l7_enviro\": true, \"l8_mission\": false}": 4,
        "GRADER-CONTRADICTION _py_first vs NONE-first": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "qwen/qwen3-vl-235b-a22b-instruct": {
      "model_slug": "qwen/qwen3-vl-235b-a22b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-coder-flash": {
      "model_slug": "qwen/qwen3-coder-flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "moonshotai/kimi-k2-0905": {
      "model_slug": "moonshotai/kimi-k2-0905",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-medium-3.1": {
      "model_slug": "mistralai/mistral-medium-3.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5-nano": {
      "model_slug": "openai/gpt-5-nano",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/codestral-2508": {
      "model_slug": "mistralai/codestral-2508",
      "cosmos_test_count": 24,
      "pass_rate_pct": 87.5,
      "twelve_cs_score": 87.6,
      "passes": 21,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=mistralai/codestral-2508": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "qwen/qwen3.5-122b-a10b": {
      "model_slug": "qwen/qwen3.5-122b-a10b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-2.0": {
      "model_slug": "aion-labs/aion-2.0",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.5-air": {
      "model_slug": "z-ai/glm-4.5-air",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-2.5-flash-lite": {
      "model_slug": "google/gemini-2.5-flash-lite",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3.5-397b-a17b": {
      "model_slug": "qwen/qwen3.5-397b-a17b",
      "cosmos_test_count": 9,
      "pass_rate_pct": 66.7,
      "twelve_cs_score": 67.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "anthropic/claude-opus-4.6": {
      "model_slug": "anthropic/claude-opus-4.6",
      "cosmos_test_count": 30,
      "pass_rate_pct": 86.7,
      "twelve_cs_score": 86.9,
      "passes": 26,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"type\\\":\\\"error\\\",\\\"error\\\":{\\\"type\\\":\\\"in": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "upstage/solar-pro-3": {
      "model_slug": "upstage/solar-pro-3",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-audio-mini": {
      "model_slug": "openai/gpt-audio-mini",
      "cosmos_test_count": 15,
      "pass_rate_pct": 20.0,
      "twelve_cs_score": 23.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 12
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "bytedance-seed/seed-1.6": {
      "model_slug": "bytedance-seed/seed-1.6",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "nvidia/nemotron-3-nano-30b-a3b": {
      "model_slug": "nvidia/nemotron-3-nano-30b-a3b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/devstral-2512": {
      "model_slug": "mistralai/devstral-2512",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "tencent/hunyuan-a13b-instruct": {
      "model_slug": "tencent/hunyuan-a13b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-small-3.2-24b-instruct": {
      "model_slug": "mistralai/mistral-small-3.2-24b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5.1-codex-max": {
      "model_slug": "openai/gpt-5.1-codex-max",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/ministral-3b-2512": {
      "model_slug": "mistralai/ministral-3b-2512",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-3-pro-image-preview": {
      "model_slug": "google/gemini-3-pro-image-preview",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "moonshotai/kimi-k2-thinking": {
      "model_slug": "moonshotai/kimi-k2-thinking",
      "cosmos_test_count": 26,
      "pass_rate_pct": 73.1,
      "twelve_cs_score": 73.9,
      "passes": 19,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "print(\"NONE\")": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"[{\\n  \\\"error\\\": {\\n    ": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-oss-safeguard-20b": {
      "model_slug": "openai/gpt-oss-safeguard-20b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5-image-mini": {
      "model_slug": "openai/gpt-5-image-mini",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/o3-pro": {
      "model_slug": "openai/o3-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-medium-3": {
      "model_slug": "mistralai/mistral-medium-3",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-14b": {
      "model_slug": "qwen/qwen3-14b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5-image": {
      "model_slug": "openai/gpt-5-image",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FORM_MISS": 9
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/o3": {
      "model_slug": "openai/o3",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4.1-nano": {
      "model_slug": "openai/gpt-4.1-nano",
      "cosmos_test_count": 26,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "write_path = \"/absolute/write/path/your_file.py\"": 4,
        "write_tool(\"/absolute/path/to/output.txt\", \"NONE\\n1\\n2\\nFizz\\n4\\nBuzz\\nFizz\\n7\\n": 4,
        "TOOL-MOUTH: replies via Write tool, not chat text": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/o1-pro": {
      "model_slug": "openai/o1-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "cohere/command-a": {
      "model_slug": "cohere/command-a",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-5-pro": {
      "model_slug": "openai/gpt-5-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "thedrummer/cydonia-24b-v4.1": {
      "model_slug": "thedrummer/cydonia-24b-v4.1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-max": {
      "model_slug": "qwen/qwen3-max",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "perplexity/sonar-reasoning-pro": {
      "model_slug": "perplexity/sonar-reasoning-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/o3-mini-high": {
      "model_slug": "openai/o3-mini-high",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-next-80b-a3b-thinking": {
      "model_slug": "qwen/qwen3-next-80b-a3b-thinking",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/o3-mini": {
      "model_slug": "openai/o3-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-01": {
      "model_slug": "minimax/minimax-01",
      "cosmos_test_count": 22,
      "pass_rate_pct": 68.2,
      "twelve_cs_score": 69.3,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "To execute the intent of printing the FizzBuzz results for numbers 1 to 15 with ": 4,
        "To fulfill the intent of printing the FizzBuzz sequence from 1 to 15 with the sp": 4,
        "NO-TOOL-ENDPOINTS": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/o1": {
      "model_slug": "openai/o1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "amazon/nova-micro-v1": {
      "model_slug": "amazon/nova-micro-v1",
      "cosmos_test_count": 26,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 26,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NONE  ": 4,
        "stop": 4,
        "form-deviation": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen-2.5-coder-32b-instruct": {
      "model_slug": "qwen/qwen-2.5-coder-32b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-3.2-1b-instruct": {
      "model_slug": "meta-llama/llama-3.2-1b-instruct",
      "cosmos_test_count": 23,
      "pass_rate_pct": 65.2,
      "twelve_cs_score": 66.4,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK llama-3.2-1b-instruct": 8,
        "\u001b[0m": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-30b-a3b-thinking-2507": {
      "model_slug": "qwen/qwen3-30b-a3b-thinking-2507",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "z-ai/glm-4.5v": {
      "model_slug": "z-ai/glm-4.5v",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-oss-120b": {
      "model_slug": "openai/gpt-oss-120b",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 8,
        "sku=openai/gpt-oss-120b": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-coder-30b-a3b-instruct": {
      "model_slug": "qwen/qwen3-coder-30b-a3b-instruct",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-235b-a22b-thinking-2507": {
      "model_slug": "qwen/qwen3-235b-a22b-thinking-2507",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-235b-a22b-2507": {
      "model_slug": "qwen/qwen3-235b-a22b-2507",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "morph/morph-v3-large": {
      "model_slug": "morph/morph-v3-large",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "REJECTED": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "cohere/command-r-plus-08-2024": {
      "model_slug": "cohere/command-r-plus-08-2024",
      "cosmos_test_count": 30,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 30,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK command-r-plus-08-2024": 8,
        "import pandas as pd": 4,
        "```python": 4,
        "form-deviation": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "sao10k/l3-lunaris-8b": {
      "model_slug": "sao10k/l3-lunaris-8b",
      "cosmos_test_count": 22,
      "pass_rate_pct": 68.2,
      "twelve_cs_score": 69.3,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "stop": 4,
        "\u001b[0m": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "minimax/minimax-m1": {
      "model_slug": "minimax/minimax-m1",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 18
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-nemo": {
      "model_slug": "mistralai/mistral-nemo",
      "cosmos_test_count": 34,
      "pass_rate_pct": 91.2,
      "twelve_cs_score": 91.1,
      "passes": 31,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "stop": 4,
        "Hello! How can I help you with your software engineering tasks today?": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o": {
      "model_slug": "openai/gpt-4o",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "google/gemini-2.5-pro-preview": {
      "model_slug": "google/gemini-2.5-pro-preview",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "SKU_UNBOUND": 6
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4-turbo": {
      "model_slug": "openai/gpt-4-turbo",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-guard-4-12b": {
      "model_slug": "meta-llama/llama-guard-4-12b",
      "cosmos_test_count": 15,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-3.5-turbo-16k": {
      "model_slug": "openai/gpt-3.5-turbo-16k",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-3.5-turbo": {
      "model_slug": "openai/gpt-3.5-turbo",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "qwen/qwen3-32b": {
      "model_slug": "qwen/qwen3-32b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/o4-mini": {
      "model_slug": "openai/o4-mini",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-4-maverick": {
      "model_slug": "meta-llama/llama-4-maverick",
      "cosmos_test_count": 45,
      "pass_rate_pct": 86.7,
      "twelve_cs_score": 86.9,
      "passes": 39,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-lines=16 first='1'": 8,
        "SLUG_DEAD": 6,
        "sku=meta-llama/llama-4-maverick": 4,
        "fizz-lines=17 first='NONE'": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-small-3.1-24b-instruct": {
      "model_slug": "mistralai/mistral-small-3.1-24b-instruct",
      "cosmos_test_count": 26,
      "pass_rate_pct": 57.7,
      "twelve_cs_score": 59.3,
      "passes": 15,
      "fails": 8,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "The intent is to create a blank file named `NONE` in the specified write_path di": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"errors\\\":[{\\\"message\\\":\\\"AiError: AiError: {\\\\\\\"error\\\\\\\":{\\\\\\\"message\\": 4,
        "{\"error\":{\"message\":\"Provider returned error\",\"code\":400,\"metadata\":{\"raw\":\"{\\\"errors\\\":[{\\\"message\\": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "rekaai/reka-flash-3": {
      "model_slug": "rekaai/reka-flash-3",
      "cosmos_test_count": 22,
      "pass_rate_pct": 68.2,
      "twelve_cs_score": 69.3,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK reka-flash-3": 8,
        "NONE  ": 4,
        "\u001b[0m": 4,
        "EMPTY": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "perplexity/sonar-pro": {
      "model_slug": "perplexity/sonar-pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "aion-labs/aion-rp-llama-3.1-8b": {
      "model_slug": "aion-labs/aion-rp-llama-3.1-8b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-small-24b-instruct-2501": {
      "model_slug": "mistralai/mistral-small-24b-instruct-2501",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "microsoft/phi-4": {
      "model_slug": "microsoft/phi-4",
      "cosmos_test_count": 18,
      "pass_rate_pct": 83.3,
      "twelve_cs_score": 83.6,
      "passes": 15,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```python": 4,
        "1": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "cohere/command-r7b-12-2024": {
      "model_slug": "cohere/command-r7b-12-2024",
      "cosmos_test_count": 18,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 18,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "```python": 4,
        "1 2 Fizz 4 Buzz Fizz 6 7 Fizz 9 Buzz Fizz 11 12 Fizz 14 Buzz Fizz 15": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "amazon/nova-pro-v1": {
      "model_slug": "amazon/nova-pro-v1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "thedrummer/unslopnemo-12b": {
      "model_slug": "thedrummer/unslopnemo-12b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "meta-llama/llama-3.2-3b-instruct": {
      "model_slug": "meta-llama/llama-3.2-3b-instruct",
      "cosmos_test_count": 22,
      "pass_rate_pct": 68.2,
      "twelve_cs_score": 69.3,
      "passes": 15,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "import os": 4,
        "stop": 4,
        "NO-TOOL-ENDPOINTS": 4,
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "sao10k/l3.1-euryale-70b": {
      "model_slug": "sao10k/l3.1-euryale-70b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o-2024-08-06": {
      "model_slug": "openai/gpt-4o-2024-08-06",
      "cosmos_test_count": 6,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o-mini": {
      "model_slug": "openai/gpt-4o-mini",
      "cosmos_test_count": 38,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 38,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HERO_OK gpt-4o-mini": 8,
        "NONE  ": 4,
        "form-deviation": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "roster",
        "live_ccr",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4o-2024-05-13": {
      "model_slug": "openai/gpt-4o-2024-05-13",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mistralai/mistral-large": {
      "model_slug": "mistralai/mistral-large",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3,
        "UPSTREAM_POOL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "mancer/weaver": {
      "model_slug": "mancer/weaver",
      "cosmos_test_count": 15,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 15
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "openai/gpt-4": {
      "model_slug": "openai/gpt-4",
      "cosmos_test_count": 6,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "HOST_FAIL": 3
      },
      "source_clusters": [
        "seats",
        "g47_seat_logs"
      ]
    },
    "stealth/space-bunny-alpha": {
      "model_slug": "stealth/space-bunny-alpha",
      "cosmos_test_count": 40,
      "pass_rate_pct": 62.5,
      "twelve_cs_score": 63.9,
      "passes": 25,
      "fails": 12,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "lines=1 first=\"http=200 error='Provider returned an emp\"": 8,
        "<empty>": 4,
        "{\"l1_role\": false, \"l2_model\": true, \"l3_harness\": true, \"l4_wrapper\": true, \"l5_skills\": true, \"l6_tools\": true, \"l7_enviro\": false, \"l8_mission\": false}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3.5-lightning:free": {
      "model_slug": "nvidia/nemotron-3.5-lightning:free",
      "cosmos_test_count": 159,
      "pass_rate_pct": 19.5,
      "twelve_cs_score": 23.0,
      "passes": 31,
      "fails": 8,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 6,
        "Here deep of. recommended formellsells Chern fells furells order: made theells20": 4,
        "Here": 4,
        "http=400 error='Only one of \"reasoning.effort\" and \"reasoning.max_tokens\" can be": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "poolside/laguna-s-2.1:free": {
      "model_slug": "poolside/laguna-s-2.1:free",
      "cosmos_test_count": 34,
      "pass_rate_pct": 50.0,
      "twelve_cs_score": 52.0,
      "passes": 17,
      "fails": 8,
      "rate_limits_429": 14,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=429 error='Provider returned error'": 8,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "poolside/laguna-xs-2.1:free": {
      "model_slug": "poolside/laguna-xs-2.1:free",
      "cosmos_test_count": 72,
      "pass_rate_pct": 23.6,
      "twelve_cs_score": 26.9,
      "passes": 17,
      "fails": 48,
      "rate_limits_429": 50,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=429 error='Provider returned error'": 24,
        "http-detail:http=429 error='Provider returned error'": 16,
        "revisit-later": 4,
        "http-detail:http=400 error='Only one of \"reasoning.effort\" and \"reasoning.max_tokens\" can be": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "cohere/north-mini-code:free": {
      "model_slug": "cohere/north-mini-code:free",
      "cosmos_test_count": 135,
      "pass_rate_pct": 14.8,
      "twelve_cs_score": 18.6,
      "passes": 20,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 6,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "womb_scars",
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3-ultra-550b-a55b:free": {
      "model_slug": "nvidia/nemotron-3-ultra-550b-a55b:free",
      "cosmos_test_count": 29,
      "pass_rate_pct": 48.3,
      "twelve_cs_score": 50.4,
      "passes": 14,
      "fails": 3,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nvidia/nemotron-3-super-120b-a12b:free": {
      "model_slug": "nvidia/nemotron-3-super-120b-a12b:free",
      "cosmos_test_count": 24,
      "pass_rate_pct": 70.8,
      "twelve_cs_score": 71.8,
      "passes": 17,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=200 error='Upstream error from Nvidia: Service temporarily overloaded'": 4,
        "fizz-16lines-ok lines=16": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "nex-agi/nex-n2.5-mini:free": {
      "model_slug": "nex-agi/nex-n2.5-mini:free",
      "cosmos_test_count": 151,
      "pass_rate_pct": 4.0,
      "twelve_cs_score": 8.3,
      "passes": 6,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=404 error='This model is unavailable for free. The paid version is availabl": 20,
        "NO_CONTEXT": 6,
        "revisit-later": 4,
        "{\"error\":{\"message\":\"This model is unavailable for free. The paid version is available now - use this slug instead: nex-agi/nex-n2.5-mini\",\"code\":404}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "seats",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "pack": {
      "model_slug": "pack",
      "cosmos_test_count": 72,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 31,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 31,
        "SLUG_DEAD": 22,
        "REJECTED": 11,
        "EMPTY": 6
      },
      "source_clusters": [
        "pack"
      ]
    },
    "codex": {
      "model_slug": "codex",
      "cosmos_test_count": 148,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "codex"
      ]
    },
    "luna6": {
      "model_slug": "luna6",
      "cosmos_test_count": 190,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "luna6"
      ]
    },
    "thinkingmachines-inkling-small-free": {
      "model_slug": "thinkingmachines-inkling-small-free",
      "cosmos_test_count": 51,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 2,
        "HARNESS_GATE": 1
      },
      "source_clusters": [
        "thinkingmachines_inkling-small_free"
      ]
    },
    "thinkingmachines-inkling-free": {
      "model_slug": "thinkingmachines-inkling-free",
      "cosmos_test_count": 25,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 2,
        "HARNESS_GATE": 1
      },
      "source_clusters": [
        "thinkingmachines_inkling_free"
      ]
    },
    "rolled": {
      "model_slug": "rolled",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "rolled"
      ]
    },
    "tampered": {
      "model_slug": "tampered",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "tampered"
      ]
    },
    "transcripts": {
      "model_slug": "transcripts",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "transcripts"
      ]
    },
    "dots-studio/dots3-note-preview:free": {
      "model_slug": "dots-studio/dots3-note-preview:free",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 6,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "openai/gpt-6-luna:floor": {
      "model_slug": "openai/gpt-6-luna:floor",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 9,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "openai/gpt-6-luna-pro:floor": {
      "model_slug": "openai/gpt-6-luna-pro:floor",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 6,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "qwen/qwen3.8-27b:free:floor": {
      "model_slug": "qwen/qwen3.8-27b:free:floor",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 3,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "google/gemma-4-26b-a4b-it:free:floor": {
      "model_slug": "google/gemma-4-26b-a4b-it:free:floor",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 3,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "inclusionai/ling-3.0-flash-vl:free": {
      "model_slug": "inclusionai/ling-3.0-flash-vl:free",
      "cosmos_test_count": 31,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 31,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=404 error='This model is unavailable for free. The paid version is availabl": 20,
        "revisit-later": 4,
        "{\"error\":{\"message\":\"This model is unavailable for free. The paid version is available now - use this slug instead: inclusionai/ling-3.0-flash-vl\",\"co": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "xiaomi/mimo-v2.5:floor": {
      "model_slug": "xiaomi/mimo-v2.5:floor",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 3,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "nex-agi/nex-n2.5-pro:free": {
      "model_slug": "nex-agi/nex-n2.5-pro:free",
      "cosmos_test_count": 34,
      "pass_rate_pct": 8.8,
      "twelve_cs_score": 12.9,
      "passes": 3,
      "fails": 28,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "http=404 error='This model is unavailable for free. The paid version is availabl": 20,
        "revisit-later": 4,
        "{\"error\":{\"message\":\"This model is unavailable for free. The paid version is available now - use this slug instead: nex-agi/nex-n2.5-pro\",\"code\":404},": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "ccr",
        "live_ccr",
        "roster"
      ]
    },
    "google/gemma-4-31b": {
      "model_slug": "google/gemma-4-31b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'g26',-'model':-'google/gemma-4-26b-a4b-it:free'}": {
      "model_slug": "{'seat':-'g26',-'model':-'google/gemma-4-26b-a4b-it:free'}",
      "cosmos_test_count": 132,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'g26p',-'model':-'google/gemma-4-26b-a4b-it'}": {
      "model_slug": "{'seat':-'g26p',-'model':-'google/gemma-4-26b-a4b-it'}",
      "cosmos_test_count": 36,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "ccr": {
      "model_slug": "ccr",
      "cosmos_test_count": 1282,
      "pass_rate_pct": 0.2,
      "twelve_cs_score": 4.7,
      "passes": 2,
      "fails": 340,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr"
      ]
    },
    "{'seat':-'llama33',-'model':-'meta-llama/llama-3.3-70b-instruct'}": {
      "model_slug": "{'seat':-'llama33',-'model':-'meta-llama/llama-3.3-70b-instruct'}",
      "cosmos_test_count": 372,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'minist',-'model':-'mistralai/ministral-8b-2512'}": {
      "model_slug": "{'seat':-'minist',-'model':-'mistralai/ministral-8b-2512'}",
      "cosmos_test_count": 459,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'ds',-'model':-'deepseek/deepseek-v4-flash-0731'}": {
      "model_slug": "{'seat':-'ds',-'model':-'deepseek/deepseek-v4-flash-0731'}",
      "cosmos_test_count": 378,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'qvl8',-'model':-'qwen/qwen3-vl-8b-instruct'}": {
      "model_slug": "{'seat':-'qvl8',-'model':-'qwen/qwen3-vl-8b-instruct'}",
      "cosmos_test_count": 450,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'ling',-'model':-'inclusionai/ling-3.0-flash-vl:free'}": {
      "model_slug": "{'seat':-'ling',-'model':-'inclusionai/ling-3.0-flash-vl:free'}",
      "cosmos_test_count": 156,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'n35f',-'model':-'nvidia/nemotron-3.5-lightning:free'}": {
      "model_slug": "{'seat':-'n35f',-'model':-'nvidia/nemotron-3.5-lightning:free'}",
      "cosmos_test_count": 348,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'seed',-'model':-'bytedance-seed/seed-2.0-mini'}": {
      "model_slug": "{'seat':-'seed',-'model':-'bytedance-seed/seed-2.0-mini'}",
      "cosmos_test_count": 84,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'qwen8',-'model':-'qwen/qwen3-8b'}": {
      "model_slug": "{'seat':-'qwen8',-'model':-'qwen/qwen3-8b'}",
      "cosmos_test_count": 81,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'qvl32',-'model':-'qwen/qwen3-vl-32b-instruct'}": {
      "model_slug": "{'seat':-'qvl32',-'model':-'qwen/qwen3-vl-32b-instruct'}",
      "cosmos_test_count": 72,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'qwen30',-'model':-'qwen/qwen3-30b-a3b'}": {
      "model_slug": "{'seat':-'qwen30',-'model':-'qwen/qwen3-30b-a3b'}",
      "cosmos_test_count": 84,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'mimo',-'model':-'xiaomi/mimo-v2.5'}": {
      "model_slug": "{'seat':-'mimo',-'model':-'xiaomi/mimo-v2.5'}",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'cursor-solo',-'model':-'cursor-cloud'}": {
      "model_slug": "{'seat':-'cursor-solo',-'model':-'cursor-cloud'}",
      "cosmos_test_count": 1059,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "{'seat':-'cursor-a',-'model':-'cursor-cloud'}": {
      "model_slug": "{'seat':-'cursor-a',-'model':-'cursor-cloud'}",
      "cosmos_test_count": 234,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "unassigned": {
      "model_slug": "unassigned",
      "cosmos_test_count": 2964,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 30
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "gemini-3.8-flash-\u2014-google-gemini-3.8-flash,-the-language-model-that-writes-the-patch.-called-through-kelly-vertex.-thinking-=-medium,-never-high.": {
      "model_slug": "gemini-3.8-flash-\u2014-google-gemini-3.8-flash,-the-language-model-that-writes-the-patch.-called-through-kelly-vertex.-thinking-=-medium,-never-high.",
      "cosmos_test_count": 12,
      "pass_rate_pct": 25.0,
      "twelve_cs_score": 28.2,
      "passes": 3,
      "fails": 6,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FAILED": 6
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "grok-4.6-\u2014-xai-grok-4.6,-the-language-model-that-writes-the-patch.-called-through-cursor-gitur-build.-not-grok.exe.": {
      "model_slug": "grok-4.6-\u2014-xai-grok-4.6,-the-language-model-that-writes-the-patch.-called-through-cursor-gitur-build.-not-grok.exe.",
      "cosmos_test_count": 897,
      "pass_rate_pct": 8.0,
      "twelve_cs_score": 12.1,
      "passes": 72,
      "fails": 279,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 138,
        "FAILED": 6
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "n/a": {
      "model_slug": "n/a",
      "cosmos_test_count": 672,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 672,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FAILED": 672
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "gpt-5.6-luna-\u2014-openai-gpt-5.6-luna,-grades-or-answers-research.-called-through-codex-cli.-first-line-item-or-none.": {
      "model_slug": "gpt-5.6-luna-\u2014-openai-gpt-5.6-luna,-grades-or-answers-research.-called-through-codex-cli.-first-line-item-or-none.",
      "cosmos_test_count": 75,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 6,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "glm-5.3-flash-\u2014-z.ai-glm-5.3-flash,-the-language-model-that-writes-the-patch.-called-through-pi-cli-(z.ai-documented-terminal-harness),-fallback-openrouter.-not-grok.exe.": {
      "model_slug": "glm-5.3-flash-\u2014-z.ai-glm-5.3-flash,-the-language-model-that-writes-the-patch.-called-through-pi-cli-(z.ai-documented-terminal-harness),-fallback-openrouter.-not-grok.exe.",
      "cosmos_test_count": 312,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "coder2": {
      "model_slug": "coder2",
      "cosmos_test_count": 9,
      "pass_rate_pct": 33.3,
      "twelve_cs_score": 36.1,
      "passes": 3,
      "fails": 6,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FAILED": 6
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "orc": {
      "model_slug": "orc",
      "cosmos_test_count": 315,
      "pass_rate_pct": 22.9,
      "twelve_cs_score": 26.3,
      "passes": 72,
      "fails": 117,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 60,
        "FAILED": 6
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "daemon": {
      "model_slug": "daemon",
      "cosmos_test_count": 672,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 672,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FAILED": 672
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "coder6": {
      "model_slug": "coder6",
      "cosmos_test_count": 426,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 153,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 72
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "research": {
      "model_slug": "research",
      "cosmos_test_count": 78,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 15,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NO_CONTEXT": 6
      },
      "source_clusters": [
        "ccr",
        "live_ccr"
      ]
    },
    "fig-database": {
      "model_slug": "fig-database",
      "cosmos_test_count": 434,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "Fig-database"
      ]
    },
    "bonsai": {
      "model_slug": "bonsai",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bonsai"
      ]
    },
    "codestral": {
      "model_slug": "codestral",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "codestral"
      ]
    },
    "ds0731": {
      "model_slug": "ds0731",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ds0731"
      ]
    },
    "ds0423": {
      "model_slug": "ds0423",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ds0423"
      ]
    },
    "dspro": {
      "model_slug": "dspro",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "dspro"
      ]
    },
    "ds41flash": {
      "model_slug": "ds41flash",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ds41flash"
      ]
    },
    "gemini38": {
      "model_slug": "gemini38",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemini38"
      ]
    },
    "gemini36": {
      "model_slug": "gemini36",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemini36"
      ]
    },
    "gemma26b": {
      "model_slug": "gemma26b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma26b"
      ]
    },
    "gemma31b": {
      "model_slug": "gemma31b",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma31b"
      ]
    },
    "glm52": {
      "model_slug": "glm52",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "glm52"
      ]
    },
    "glm53": {
      "model_slug": "glm53",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "glm53"
      ]
    },
    "glm53flash": {
      "model_slug": "glm53flash",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "glm53flash"
      ]
    },
    "gpt6luna": {
      "model_slug": "gpt6luna",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gpt6luna"
      ]
    },
    "gpt6lunapro": {
      "model_slug": "gpt6lunapro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gpt6lunapro"
      ]
    },
    "grok46": {
      "model_slug": "grok46",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "grok46"
      ]
    },
    "hy3": {
      "model_slug": "hy3",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hy3"
      ]
    },
    "lagunas": {
      "model_slug": "lagunas",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lagunas"
      ]
    },
    "inkling": {
      "model_slug": "inkling",
      "cosmos_test_count": 36,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "inkling"
      ]
    },
    "lagunaxs": {
      "model_slug": "lagunaxs",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lagunaxs"
      ]
    },
    "lingflash": {
      "model_slug": "lingflash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lingflash"
      ]
    },
    "lingflashvl": {
      "model_slug": "lingflashvl",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lingflashvl"
      ]
    },
    "llama3370b": {
      "model_slug": "llama3370b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "llama3370b"
      ]
    },
    "luna": {
      "model_slug": "luna",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "luna"
      ]
    },
    "lunapro": {
      "model_slug": "lunapro",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lunapro"
      ]
    },
    "maverick": {
      "model_slug": "maverick",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "maverick"
      ]
    },
    "mimo25": {
      "model_slug": "mimo25",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "mimo25"
      ]
    },
    "ministral": {
      "model_slug": "ministral",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ministral"
      ]
    },
    "muse12": {
      "model_slug": "muse12",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "muse12"
      ]
    },
    "muse11c": {
      "model_slug": "muse11c",
      "cosmos_test_count": 14,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "muse11c"
      ]
    },
    "muse12c": {
      "model_slug": "muse12c",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "muse12c"
      ]
    },
    "nemolight": {
      "model_slug": "nemolight",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nemolight"
      ]
    },
    "muse13c": {
      "model_slug": "muse13c",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "muse13c"
      ]
    },
    "nemoultra": {
      "model_slug": "nemoultra",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nemoultra"
      ]
    },
    "oss120b": {
      "model_slug": "oss120b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "oss120b"
      ]
    },
    "oss20b": {
      "model_slug": "oss20b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "oss20b"
      ]
    },
    "qwen330b": {
      "model_slug": "qwen330b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen330b"
      ]
    },
    "qwen359b": {
      "model_slug": "qwen359b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen359b"
      ]
    },
    "qwen3635b": {
      "model_slug": "qwen3635b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen3635b"
      ]
    },
    "qwen3827b": {
      "model_slug": "qwen3827b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen3827b"
      ]
    },
    "qwen37flash": {
      "model_slug": "qwen37flash",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen37flash"
      ]
    },
    "qwen38b": {
      "model_slug": "qwen38b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen38b"
      ]
    },
    "qwen38flash": {
      "model_slug": "qwen38flash",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen38flash"
      ]
    },
    "qwen38max": {
      "model_slug": "qwen38max",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen38max"
      ]
    },
    "qwen3vl32b": {
      "model_slug": "qwen3vl32b",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen3vl32b"
      ]
    },
    "qwen3vl8b": {
      "model_slug": "qwen3vl8b",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "qwen3vl8b"
      ]
    },
    "seed20mini": {
      "model_slug": "seed20mini",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seed20mini"
      ]
    },
    "scout": {
      "model_slug": "scout",
      "cosmos_test_count": 14,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "scout"
      ]
    },
    "sol": {
      "model_slug": "sol",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "sol"
      ]
    },
    "solarmini4": {
      "model_slug": "solarmini4",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "solarmini4"
      ]
    },
    "solarpro4": {
      "model_slug": "solarpro4",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "solarpro4"
      ]
    },
    "terra": {
      "model_slug": "terra",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "terra"
      ]
    },
    "f2-lagunasf": {
      "model_slug": "f2-lagunasf",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "F2_lagunasf"
      ]
    },
    "f2-lagunaxsf-fizz": {
      "model_slug": "f2-lagunaxsf-fizz",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 5,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "F2_lagunaxsf_fizz"
      ]
    },
    "f2-qwen27f-fizz": {
      "model_slug": "f2-qwen27f-fizz",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 5,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "F2_qwen27f_fizz"
      ]
    },
    "fc-gemma31bf-fizz": {
      "model_slug": "fc-gemma31bf-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 4,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma31bf_fizz"
      ]
    },
    "fc-gemma26bf-fizz2": {
      "model_slug": "fc-gemma26bf-fizz2",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma26bf_fizz2"
      ]
    },
    "fc-gemma26bf-task": {
      "model_slug": "fc-gemma26bf-task",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 4,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma26bf_task"
      ]
    },
    "fc-gemma26bf-fizz": {
      "model_slug": "fc-gemma26bf-fizz",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 4,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma26bf_fizz"
      ]
    },
    "fc-gemma31bf-fizz2": {
      "model_slug": "fc-gemma31bf-fizz2",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma31bf_fizz2"
      ]
    },
    "fc-gemma31bf-task": {
      "model_slug": "fc-gemma31bf-task",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 4,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_gemma31bf_task"
      ]
    },
    "observed-usage": {
      "model_slug": "observed-usage",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "observed-usage"
      ]
    },
    "fc-lagunaxsf-fizz": {
      "model_slug": "fc-lagunaxsf-fizz",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 7,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_lagunaxsf_fizz"
      ]
    },
    "fc-lagunaxsf-task": {
      "model_slug": "fc-lagunaxsf-task",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 8,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_lagunaxsf_task"
      ]
    },
    "fc-nexmini-paid": {
      "model_slug": "fc-nexmini-paid",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_nexmini_paid"
      ]
    },
    "fc-nexpro-paid": {
      "model_slug": "fc-nexpro-paid",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "FC_nexpro_paid"
      ]
    },
    "gemma26bf": {
      "model_slug": "gemma26bf",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma26bf"
      ]
    },
    "gemma26bf-fizz": {
      "model_slug": "gemma26bf-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 5,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma26bf_fizz"
      ]
    },
    "gemma31bf": {
      "model_slug": "gemma31bf",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma31bf"
      ]
    },
    "gemma31bf-fizz": {
      "model_slug": "gemma31bf-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 5,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "gemma31bf_fizz"
      ]
    },
    "go9-dots3": {
      "model_slug": "go9-dots3",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_dots3"
      ]
    },
    "go9-gemma26bf": {
      "model_slug": "go9-gemma26bf",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_gemma26bf"
      ]
    },
    "go9-gemma31bf": {
      "model_slug": "go9-gemma31bf",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_gemma31bf"
      ]
    },
    "go9-hy3": {
      "model_slug": "go9-hy3",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_hy3"
      ]
    },
    "go9-nanoomni": {
      "model_slug": "go9-nanoomni",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_nanoomni"
      ]
    },
    "go9-maverick": {
      "model_slug": "go9-maverick",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_maverick"
      ]
    },
    "go9-qwen27f": {
      "model_slug": "go9-qwen27f",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 7,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_qwen27f"
      ]
    },
    "go9-qwen330b": {
      "model_slug": "go9-qwen330b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_qwen330b"
      ]
    },
    "go9-qwen38b": {
      "model_slug": "go9-qwen38b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "GO9_qwen38b"
      ]
    },
    "inklingf-fizz": {
      "model_slug": "inklingf-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "inklingf_fizz"
      ]
    },
    "inklingsm": {
      "model_slug": "inklingsm",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "inklingsm"
      ]
    },
    "lagunaxsf": {
      "model_slug": "lagunaxsf",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lagunaxsf"
      ]
    },
    "inklingsm-fizz": {
      "model_slug": "inklingsm-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "inklingsm_fizz"
      ]
    },
    "lightning-fizz": {
      "model_slug": "lightning-fizz",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lightning_fizz"
      ]
    },
    "new-commandaplus": {
      "model_slug": "new-commandaplus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_commandaplus"
      ]
    },
    "lingvlf-fizz": {
      "model_slug": "lingvlf-fizz",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lingvlf_fizz"
      ]
    },
    "new-commandaplus-fizz": {
      "model_slug": "new-commandaplus-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_commandaplus_fizz"
      ]
    },
    "new-dsv4pro0423": {
      "model_slug": "new-dsv4pro0423",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_dsv4pro0423"
      ]
    },
    "new-dsv4flashvis": {
      "model_slug": "new-dsv4flashvis",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_dsv4flashvis"
      ]
    },
    "new-glm53batch": {
      "model_slug": "new-glm53batch",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_glm53batch"
      ]
    },
    "new-dsv4flashvis-fizz": {
      "model_slug": "new-dsv4flashvis-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_dsv4flashvis_fizz"
      ]
    },
    "new-glm53batch-fizz": {
      "model_slug": "new-glm53batch-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_glm53batch_fizz"
      ]
    },
    "new-glm53flashx-fizz": {
      "model_slug": "new-glm53flashx-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_glm53flashx_fizz"
      ]
    },
    "new-glm53flashx": {
      "model_slug": "new-glm53flashx",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_glm53flashx"
      ]
    },
    "new-inklingsmp": {
      "model_slug": "new-inklingsmp",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_inklingsmp"
      ]
    },
    "new-inklingsmp-fizz": {
      "model_slug": "new-inklingsmp-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_inklingsmp_fizz"
      ]
    },
    "new-kimi26": {
      "model_slug": "new-kimi26",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_kimi26"
      ]
    },
    "new-kimi26-fizz": {
      "model_slug": "new-kimi26-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_kimi26_fizz"
      ]
    },
    "new-longcat20": {
      "model_slug": "new-longcat20",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_longcat20"
      ]
    },
    "new-longcat20-fizz": {
      "model_slug": "new-longcat20-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_longcat20_fizz"
      ]
    },
    "new-mimo25pro": {
      "model_slug": "new-mimo25pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_mimo25pro"
      ]
    },
    "new-mimo25pro-fizz": {
      "model_slug": "new-mimo25pro-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_mimo25pro_fizz"
      ]
    },
    "new-mimo26pro": {
      "model_slug": "new-mimo26pro",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_mimo26pro"
      ]
    },
    "new-mimo26pro-fizz": {
      "model_slug": "new-mimo26pro-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_mimo26pro_fizz"
      ]
    },
    "new-minimaxm3-fizz": {
      "model_slug": "new-minimaxm3-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_minimaxm3_fizz"
      ]
    },
    "new-museglimmer": {
      "model_slug": "new-museglimmer",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_museglimmer"
      ]
    },
    "new-minimaxm3": {
      "model_slug": "new-minimaxm3",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_minimaxm3"
      ]
    },
    "new-museglimmer-fizz": {
      "model_slug": "new-museglimmer-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_museglimmer_fizz"
      ]
    },
    "new-perceptronmk1": {
      "model_slug": "new-perceptronmk1",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_perceptronmk1"
      ]
    },
    "new-perceptronmk15-fizz": {
      "model_slug": "new-perceptronmk15-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_perceptronmk15_fizz"
      ]
    },
    "new-perceptronmk15": {
      "model_slug": "new-perceptronmk15",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_perceptronmk15"
      ]
    },
    "new-perceptronmk1-fizz": {
      "model_slug": "new-perceptronmk1-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_perceptronmk1_fizz"
      ]
    },
    "new-qwen35plus": {
      "model_slug": "new-qwen35plus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen35plus"
      ]
    },
    "new-qwen3627b": {
      "model_slug": "new-qwen3627b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen3627b"
      ]
    },
    "new-qwen35plus-fizz": {
      "model_slug": "new-qwen35plus-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen35plus_fizz"
      ]
    },
    "new-qwen3627b-fizz": {
      "model_slug": "new-qwen3627b-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen3627b_fizz"
      ]
    },
    "new-qwen36flash-fizz": {
      "model_slug": "new-qwen36flash-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen36flash_fizz"
      ]
    },
    "new-qwen36flash": {
      "model_slug": "new-qwen36flash",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen36flash"
      ]
    },
    "new-qwen36plus": {
      "model_slug": "new-qwen36plus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen36plus"
      ]
    },
    "new-qwen36plus-fizz": {
      "model_slug": "new-qwen36plus-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen36plus_fizz"
      ]
    },
    "new-qwen37plus": {
      "model_slug": "new-qwen37plus",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen37plus"
      ]
    },
    "new-qwen37plus-fizz": {
      "model_slug": "new-qwen37plus-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_qwen37plus_fizz"
      ]
    },
    "new-seed20code-fizz": {
      "model_slug": "new-seed20code-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_seed20code_fizz"
      ]
    },
    "new-seed21turbo": {
      "model_slug": "new-seed21turbo",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_seed21turbo"
      ]
    },
    "new-seed20code": {
      "model_slug": "new-seed20code",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_seed20code"
      ]
    },
    "new-step37flash": {
      "model_slug": "new-step37flash",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_step37flash"
      ]
    },
    "new-seed21turbo-fizz": {
      "model_slug": "new-seed21turbo-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_seed21turbo_fizz"
      ]
    },
    "new-step37flash-fizz": {
      "model_slug": "new-step37flash-fizz",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "NEW_step37flash_fizz"
      ]
    },
    "nexmini": {
      "model_slug": "nexmini",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nexmini"
      ]
    },
    "nexmini-fizz": {
      "model_slug": "nexmini-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nexmini_fizz"
      ]
    },
    "nexpro": {
      "model_slug": "nexpro",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nexpro"
      ]
    },
    "nexpro-fizz": {
      "model_slug": "nexpro-fizz",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "nexpro_fizz"
      ]
    },
    "retry2-gemma31bit": {
      "model_slug": "retry2-gemma31bit",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_gemma31bit"
      ]
    },
    "retry2-inklingp": {
      "model_slug": "retry2-inklingp",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_inklingp"
      ]
    },
    "retry2-nanoomni": {
      "model_slug": "retry2-nanoomni",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_nanoomni"
      ]
    },
    "retry2-nemolight": {
      "model_slug": "retry2-nemolight",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_nemolight"
      ]
    },
    "retry2-qwen27f": {
      "model_slug": "retry2-qwen27f",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 6,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_qwen27f"
      ]
    },
    "retry2-qwen38b": {
      "model_slug": "retry2-qwen38b",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_qwen38b"
      ]
    },
    "retry2-qwen3vl32b": {
      "model_slug": "retry2-qwen3vl32b",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_qwen3vl32b"
      ]
    },
    "retry2-qwen3vl8b": {
      "model_slug": "retry2-qwen3vl8b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_qwen3vl8b"
      ]
    },
    "retry2-seed-notools": {
      "model_slug": "retry2-seed-notools",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY2_seed-notools"
      ]
    },
    "retry-muse11bare": {
      "model_slug": "retry-muse11bare",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY_muse11bare"
      ]
    },
    "retry-inklingp": {
      "model_slug": "retry-inklingp",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY_inklingp"
      ]
    },
    "retry-qwen38b": {
      "model_slug": "retry-qwen38b",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY_qwen38b"
      ]
    },
    "retry-qwen3vl32b": {
      "model_slug": "retry-qwen3vl32b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "RETRY_qwen3vl32b"
      ]
    },
    "write-dots3": {
      "model_slug": "write-dots3",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_dots3"
      ]
    },
    "super": {
      "model_slug": "super",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "super"
      ]
    },
    "write-gemma31bit": {
      "model_slug": "write-gemma31bit",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_gemma31bit"
      ]
    },
    "write-inklingp": {
      "model_slug": "write-inklingp",
      "cosmos_test_count": 7,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_inklingp"
      ]
    },
    "write-qwen3vl32b": {
      "model_slug": "write-qwen3vl32b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_qwen3vl32b"
      ]
    },
    "write-maverick": {
      "model_slug": "write-maverick",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_maverick"
      ]
    },
    "write-qwen3vl8b": {
      "model_slug": "write-qwen3vl8b",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_qwen3vl8b"
      ]
    },
    "write-seed-notools": {
      "model_slug": "write-seed-notools",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "WRITE_seed-notools"
      ]
    },
    "meta/muse-spark-1.1-contributor": {
      "model_slug": "meta/muse-spark-1.1-contributor",
      "cosmos_test_count": 32,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 32,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "mouth-form": 28,
        "{\"error\":{\"message\":\"meta/muse-spark-1.1-contributor is not a valid model ID\",\"code\":400},\"user_id\":\"user_3Iw6zwRpqCi2CdefGLtSeuerqqT\"}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "roster",
        "live_ccr"
      ]
    },
    "z-ai/glm-5.3:batch": {
      "model_slug": "z-ai/glm-5.3:batch",
      "cosmos_test_count": 44,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 44,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'z-ai/glm-5.3:batch'; want ['bytedance-seed/seed-2.0-|PIN-NOTE": 28,
        "{\"error\":{\"message\":\"Not Found\",\"code\":404}}": 8,
        "revisit-later|PIN-NOTE": 4,
        "{\"error\":{\"message\":\"z-ai/glm-5.3:batch cannot be used with the chat/completions endpoint (adapter DeepInfraBatchAdapter": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "roster",
        "live_ccr"
      ]
    },
    "deepseek/deepseek-v4-pro-0423": {
      "model_slug": "deepseek/deepseek-v4-pro-0423",
      "cosmos_test_count": 16,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 16,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "not a pinned OpenRouter id 'deepseek/deepseek-v4-pro-0423'; want ['bytedance-see|PIN-NOTE": 8,
        "revisit-later|PIN-NOTE": 4,
        "{\"error\":{\"message\":\"deepseek/deepseek-v4-pro-0423 is not a valid model ID\",\"code\":400},\"user_id\":\"user_3Iw6zwRpqCi2CdefGLtSeuerqqT\"}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "roster",
        "live_ccr"
      ]
    },
    "qwen/qwen2.5-7b-instruct": {
      "model_slug": "qwen/qwen2.5-7b-instruct",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 4,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "{\"error\":{\"message\":\"qwen/qwen2.5-7b-instruct is not a valid model ID\",\"code\":400},\"user_id\":\"user_3Iw6zwRpqCi2CdefGLtSeuerqqT\"}": 4
      },
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "roster",
        "live_ccr"
      ]
    },
    "copilot/auto": {
      "model_slug": "copilot/auto",
      "cosmos_test_count": 8,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 8,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "01_SEATING_DATA",
        "02_SEATING_AND_ATTEMPTS",
        "roster",
        "live_ccr"
      ]
    },
    "roster": {
      "model_slug": "roster",
      "cosmos_test_count": 48,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "roster"
      ]
    },
    "goal-round-driver": {
      "model_slug": "goal-round-driver",
      "cosmos_test_count": 58,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "goal-round-driver"
      ]
    },
    "goal-wrapup": {
      "model_slug": "goal-wrapup",
      "cosmos_test_count": 51,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "goal-wrapup"
      ]
    },
    "goal-tools": {
      "model_slug": "goal-tools",
      "cosmos_test_count": 32,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "goal-tools"
      ]
    },
    "headless-profile": {
      "model_slug": "headless-profile",
      "cosmos_test_count": 24,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "headless-profile"
      ]
    },
    "invalid-credential": {
      "model_slug": "invalid-credential",
      "cosmos_test_count": 15,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "invalid-credential"
      ]
    },
    "missing-credential": {
      "model_slug": "missing-credential",
      "cosmos_test_count": 15,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "missing-credential"
      ]
    },
    "semantic-checkpoint": {
      "model_slug": "semantic-checkpoint",
      "cosmos_test_count": 28,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "semantic-checkpoint"
      ]
    },
    "provider-retry": {
      "model_slug": "provider-retry",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "provider-retry"
      ]
    },
    "subagent-diagnostic": {
      "model_slug": "subagent-diagnostic",
      "cosmos_test_count": 30,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-diagnostic"
      ]
    },
    "subagent-inheritance": {
      "model_slug": "subagent-inheritance",
      "cosmos_test_count": 85,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-inheritance"
      ]
    },
    "subagent-settlement": {
      "model_slug": "subagent-settlement",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-settlement"
      ]
    },
    "offline-edit": {
      "model_slug": "offline-edit",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "offline-edit"
      ]
    },
    "precedence-change": {
      "model_slug": "precedence-change",
      "cosmos_test_count": 26,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "precedence-change"
      ]
    },
    "preview-architecture-review": {
      "model_slug": "preview-architecture-review",
      "cosmos_test_count": 356,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "preview-architecture-review"
      ]
    },
    "preview-follow-up-builder": {
      "model_slug": "preview-follow-up-builder",
      "cosmos_test_count": 16,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "preview-follow-up-builder"
      ]
    },
    "preview-showcase": {
      "model_slug": "preview-showcase",
      "cosmos_test_count": 400,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "preview-showcase"
      ]
    },
    "fixtures": {
      "model_slug": "fixtures",
      "cosmos_test_count": 25,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fixtures"
      ]
    },
    "rec-child": {
      "model_slug": "rec-child",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "rec-child"
      ]
    },
    "rec-pin": {
      "model_slug": "rec-pin",
      "cosmos_test_count": 13,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "rec-pin"
      ]
    },
    "authored-error": {
      "model_slug": "authored-error",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "authored-error"
      ]
    },
    "blocked-log": {
      "model_slug": "blocked-log",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "blocked-log"
      ]
    },
    "pin-turn": {
      "model_slug": "pin-turn",
      "cosmos_test_count": 24,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "pin-turn"
      ]
    },
    "plain-turn": {
      "model_slug": "plain-turn",
      "cosmos_test_count": 29,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "plain-turn"
      ]
    },
    "shared-pin": {
      "model_slug": "shared-pin",
      "cosmos_test_count": 19,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "shared-pin"
      ]
    },
    "advanced": {
      "model_slug": "advanced",
      "cosmos_test_count": 335,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "advanced"
      ]
    },
    "scheduler-recovery": {
      "model_slug": "scheduler-recovery",
      "cosmos_test_count": 62,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "scheduler-recovery"
      ]
    },
    "restart": {
      "model_slug": "restart",
      "cosmos_test_count": 106,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "restart"
      ]
    },
    "cancel": {
      "model_slug": "cancel",
      "cosmos_test_count": 37,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cancel"
      ]
    },
    "cancel-tool-calls": {
      "model_slug": "cancel-tool-calls",
      "cosmos_test_count": 48,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cancel-tool-calls"
      ]
    },
    "escalation-approved": {
      "model_slug": "escalation-approved",
      "cosmos_test_count": 56,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "escalation-approved"
      ]
    },
    "escalation-rejected": {
      "model_slug": "escalation-rejected",
      "cosmos_test_count": 80,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "escalation-rejected"
      ]
    },
    "fs-escalation-approved": {
      "model_slug": "fs-escalation-approved",
      "cosmos_test_count": 56,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-escalation-approved"
      ]
    },
    "fs-same-mode": {
      "model_slug": "fs-same-mode",
      "cosmos_test_count": 30,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-same-mode"
      ]
    },
    "handshake": {
      "model_slug": "handshake",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "handshake"
      ]
    },
    "image-compaction": {
      "model_slug": "image-compaction",
      "cosmos_test_count": 99,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "image-compaction"
      ]
    },
    "reject-extra-dirs": {
      "model_slug": "reject-extra-dirs",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "reject-extra-dirs"
      ]
    },
    "account-provider-signout": {
      "model_slug": "account-provider-signout",
      "cosmos_test_count": 49,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "account-provider-signout"
      ]
    },
    "dynamic-tool-updates": {
      "model_slug": "dynamic-tool-updates",
      "cosmos_test_count": 91,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "dynamic-tool-updates"
      ]
    },
    "inline-image-prompt": {
      "model_slug": "inline-image-prompt",
      "cosmos_test_count": 36,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "inline-image-prompt"
      ]
    },
    "bash-tool": {
      "model_slug": "bash-tool",
      "cosmos_test_count": 63,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bash-tool"
      ]
    },
    "max-tokens-continue": {
      "model_slug": "max-tokens-continue",
      "cosmos_test_count": 49,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "max-tokens-continue"
      ]
    },
    "persistent-tools": {
      "model_slug": "persistent-tools",
      "cosmos_test_count": 147,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "persistent-tools"
      ]
    },
    "multi-turn": {
      "model_slug": "multi-turn",
      "cosmos_test_count": 49,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "multi-turn"
      ]
    },
    "serial-created": {
      "model_slug": "serial-created",
      "cosmos_test_count": 46,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "serial-created"
      ]
    },
    "ptc-turn": {
      "model_slug": "ptc-turn",
      "cosmos_test_count": 101,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-turn"
      ]
    },
    "session-title-after-turn": {
      "model_slug": "session-title-after-turn",
      "cosmos_test_count": 37,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "session-title-after-turn"
      ]
    },
    "subagent-activation-limit": {
      "model_slug": "subagent-activation-limit",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-activation-limit"
      ]
    },
    "subagent-continuable": {
      "model_slug": "subagent-continuable",
      "cosmos_test_count": 145,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-continuable"
      ]
    },
    "subagent-continuable-inheritance": {
      "model_slug": "subagent-continuable-inheritance",
      "cosmos_test_count": 95,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-continuable-inheritance"
      ]
    },
    "subagent-dsh-sdk-diagnostic": {
      "model_slug": "subagent-dsh-sdk-diagnostic",
      "cosmos_test_count": 105,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-dsh-sdk-diagnostic"
      ]
    },
    "subagent-dsh-sdk-dynamic-route": {
      "model_slug": "subagent-dsh-sdk-dynamic-route",
      "cosmos_test_count": 98,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-dsh-sdk-dynamic-route"
      ]
    },
    "subagent-fork-in-process": {
      "model_slug": "subagent-fork-in-process",
      "cosmos_test_count": 122,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-fork-in-process"
      ]
    },
    "subagent-mixed": {
      "model_slug": "subagent-mixed",
      "cosmos_test_count": 168,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-mixed"
      ]
    },
    "subagent-list-agents": {
      "model_slug": "subagent-list-agents",
      "cosmos_test_count": 183,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-list-agents"
      ]
    },
    "subagent-send-message": {
      "model_slug": "subagent-send-message",
      "cosmos_test_count": 127,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-send-message"
      ]
    },
    "subagent-spawn-in-process": {
      "model_slug": "subagent-spawn-in-process",
      "cosmos_test_count": 195,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-spawn-in-process"
      ]
    },
    "tool-error-details": {
      "model_slug": "tool-error-details",
      "cosmos_test_count": 85,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "tool-error-details"
      ]
    },
    "system-prompt-in-history": {
      "model_slug": "system-prompt-in-history",
      "cosmos_test_count": 67,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "system-prompt-in-history"
      ]
    },
    "text-turn": {
      "model_slug": "text-turn",
      "cosmos_test_count": 138,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "text-turn"
      ]
    },
    "tool-scheduler-recovery": {
      "model_slug": "tool-scheduler-recovery",
      "cosmos_test_count": 62,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "tool-scheduler-recovery"
      ]
    },
    "advanced-toolchain": {
      "model_slug": "advanced-toolchain",
      "cosmos_test_count": 255,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "advanced-toolchain"
      ]
    },
    "advanced-toolchain-runtime": {
      "model_slug": "advanced-toolchain-runtime",
      "cosmos_test_count": 232,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "advanced-toolchain-runtime"
      ]
    },
    "agent-instructions": {
      "model_slug": "agent-instructions",
      "cosmos_test_count": 77,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "agent-instructions"
      ]
    },
    "background-confinement-failure": {
      "model_slug": "background-confinement-failure",
      "cosmos_test_count": 30,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "background-confinement-failure"
      ]
    },
    "background-job-admission": {
      "model_slug": "background-job-admission",
      "cosmos_test_count": 73,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "background-job-admission"
      ]
    },
    "bash-same-mode-empty-justification": {
      "model_slug": "bash-same-mode-empty-justification",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bash-same-mode-empty-justification"
      ]
    },
    "bash-spill": {
      "model_slug": "bash-spill",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bash-spill"
      ]
    },
    "bash-startup-timeout": {
      "model_slug": "bash-startup-timeout",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bash-startup-timeout"
      ]
    },
    "bash-tool-turn": {
      "model_slug": "bash-tool-turn",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "bash-tool-turn"
      ]
    },
    "both-mode-turn": {
      "model_slug": "both-mode-turn",
      "cosmos_test_count": 47,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "both-mode-turn"
      ]
    },
    "browser-use-chrome-devtools-mcp": {
      "model_slug": "browser-use-chrome-devtools-mcp",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "browser-use-chrome-devtools-mcp"
      ]
    },
    "browser-use-playwright-mcp": {
      "model_slug": "browser-use-playwright-mcp",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "browser-use-playwright-mcp"
      ]
    },
    "browser-use-stagehand-native": {
      "model_slug": "browser-use-stagehand-native",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "browser-use-stagehand-native"
      ]
    },
    "compaction-output-reserve": {
      "model_slug": "compaction-output-reserve",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "compaction-output-reserve"
      ]
    },
    "compaction-recovery": {
      "model_slug": "compaction-recovery",
      "cosmos_test_count": 55,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "compaction-recovery"
      ]
    },
    "compaction-summary-headroom": {
      "model_slug": "compaction-summary-headroom",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "compaction-summary-headroom"
      ]
    },
    "computer-use-cua-driver-mcp": {
      "model_slug": "computer-use-cua-driver-mcp",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "computer-use-cua-driver-mcp"
      ]
    },
    "computer-use-cua-driver-native": {
      "model_slug": "computer-use-cua-driver-native",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "computer-use-cua-driver-native"
      ]
    },
    "cordis-inspect-jsdoc": {
      "model_slug": "cordis-inspect-jsdoc",
      "cosmos_test_count": 80,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cordis-inspect-jsdoc"
      ]
    },
    "cordis-inspect-timeout": {
      "model_slug": "cordis-inspect-timeout",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cordis-inspect-timeout"
      ]
    },
    "cordis-inspect-liveness": {
      "model_slug": "cordis-inspect-liveness",
      "cosmos_test_count": 35,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cordis-inspect-liveness"
      ]
    },
    "deepseek-messages-degraded-replay": {
      "model_slug": "deepseek-messages-degraded-replay",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-degraded-replay"
      ]
    },
    "deepseek-messages-invalid-tool-history": {
      "model_slug": "deepseek-messages-invalid-tool-history",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-invalid-tool-history"
      ]
    },
    "deepseek-messages-input-history": {
      "model_slug": "deepseek-messages-input-history",
      "cosmos_test_count": 19,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-input-history"
      ]
    },
    "deepseek-messages-replay": {
      "model_slug": "deepseek-messages-replay",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-replay"
      ]
    },
    "deepseek-messages-system-prompt": {
      "model_slug": "deepseek-messages-system-prompt",
      "cosmos_test_count": 23,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-system-prompt"
      ]
    },
    "dynamic-tool-prompt-updates": {
      "model_slug": "dynamic-tool-prompt-updates",
      "cosmos_test_count": 32,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "dynamic-tool-prompt-updates"
      ]
    },
    "deepseek-protocol-system-prompt": {
      "model_slug": "deepseek-protocol-system-prompt",
      "cosmos_test_count": 23,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-protocol-system-prompt"
      ]
    },
    "empty-response-retry": {
      "model_slug": "empty-response-retry",
      "cosmos_test_count": 45,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "empty-response-retry"
      ]
    },
    "empty-response-retry-current": {
      "model_slug": "empty-response-retry-current",
      "cosmos_test_count": 20,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "empty-response-retry-current"
      ]
    },
    "error-finish": {
      "model_slug": "error-finish",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "error-finish"
      ]
    },
    "fs-edit": {
      "model_slug": "fs-edit",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-edit"
      ]
    },
    "fs-delete-recreate": {
      "model_slug": "fs-delete-recreate",
      "cosmos_test_count": 73,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-delete-recreate"
      ]
    },
    "foreground-confinement-timeout": {
      "model_slug": "foreground-confinement-timeout",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "foreground-confinement-timeout"
      ]
    },
    "fs-glob-sampling": {
      "model_slug": "fs-glob-sampling",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-glob-sampling"
      ]
    },
    "fs-policy-reject": {
      "model_slug": "fs-policy-reject",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-policy-reject"
      ]
    },
    "fs-read": {
      "model_slug": "fs-read",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-read"
      ]
    },
    "fs-read-window": {
      "model_slug": "fs-read-window",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-read-window"
      ]
    },
    "fs-write": {
      "model_slug": "fs-write",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-write"
      ]
    },
    "fs-write-overwrite-bounded": {
      "model_slug": "fs-write-overwrite-bounded",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-write-overwrite-bounded"
      ]
    },
    "fs-write-overwrite": {
      "model_slug": "fs-write-overwrite",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fs-write-overwrite"
      ]
    },
    "hook-cc-invalid-matcher": {
      "model_slug": "hook-cc-invalid-matcher",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-invalid-matcher"
      ]
    },
    "hook-cc-posttool-block": {
      "model_slug": "hook-cc-posttool-block",
      "cosmos_test_count": 61,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-posttool-block"
      ]
    },
    "hook-cc-posttool-context": {
      "model_slug": "hook-cc-posttool-context",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-posttool-context"
      ]
    },
    "hook-cc-pretool-ask": {
      "model_slug": "hook-cc-pretool-ask",
      "cosmos_test_count": 51,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-pretool-ask"
      ]
    },
    "hook-cc-pretool-deny": {
      "model_slug": "hook-cc-pretool-deny",
      "cosmos_test_count": 47,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-pretool-deny"
      ]
    },
    "hook-cc-promptsubmit-block": {
      "model_slug": "hook-cc-promptsubmit-block",
      "cosmos_test_count": 20,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-promptsubmit-block"
      ]
    },
    "hook-cc-promptsubmit-context": {
      "model_slug": "hook-cc-promptsubmit-context",
      "cosmos_test_count": 39,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-promptsubmit-context"
      ]
    },
    "hook-cc-stop-continue": {
      "model_slug": "hook-cc-stop-continue",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-cc-stop-continue"
      ]
    },
    "hook-codex-invalid-matcher": {
      "model_slug": "hook-codex-invalid-matcher",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-invalid-matcher"
      ]
    },
    "hook-codex-posttool-block": {
      "model_slug": "hook-codex-posttool-block",
      "cosmos_test_count": 47,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-posttool-block"
      ]
    },
    "hook-codex-posttool-context": {
      "model_slug": "hook-codex-posttool-context",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-posttool-context"
      ]
    },
    "hook-codex-pretool-block": {
      "model_slug": "hook-codex-pretool-block",
      "cosmos_test_count": 47,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-pretool-block"
      ]
    },
    "hook-codex-promptsubmit-block": {
      "model_slug": "hook-codex-promptsubmit-block",
      "cosmos_test_count": 20,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-promptsubmit-block"
      ]
    },
    "hook-codex-promptsubmit-context": {
      "model_slug": "hook-codex-promptsubmit-context",
      "cosmos_test_count": 39,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-promptsubmit-context"
      ]
    },
    "hook-codex-stop-continue": {
      "model_slug": "hook-codex-stop-continue",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "hook-codex-stop-continue"
      ]
    },
    "lsp-definition": {
      "model_slug": "lsp-definition",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lsp-definition"
      ]
    },
    "mcp-empty": {
      "model_slug": "mcp-empty",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "mcp-empty"
      ]
    },
    "mcp-empty-ptc": {
      "model_slug": "mcp-empty-ptc",
      "cosmos_test_count": 26,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "mcp-empty-ptc"
      ]
    },
    "mcp-resources": {
      "model_slug": "mcp-resources",
      "cosmos_test_count": 36,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "mcp-resources"
      ]
    },
    "mcp-resources-ptc": {
      "model_slug": "mcp-resources-ptc",
      "cosmos_test_count": 32,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "mcp-resources-ptc"
      ]
    },
    "missing-sandbox-runner": {
      "model_slug": "missing-sandbox-runner",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "missing-sandbox-runner"
      ]
    },
    "model-switch-notice": {
      "model_slug": "model-switch-notice",
      "cosmos_test_count": 52,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "model-switch-notice"
      ]
    },
    "multimodal-spill-ends": {
      "model_slug": "multimodal-spill-ends",
      "cosmos_test_count": 44,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "multimodal-spill-ends"
      ]
    },
    "multimodal-spill-middle": {
      "model_slug": "multimodal-spill-middle",
      "cosmos_test_count": 44,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "multimodal-spill-middle"
      ]
    },
    "office-skills": {
      "model_slug": "office-skills",
      "cosmos_test_count": 50,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "office-skills"
      ]
    },
    "office-skills-no-renderer": {
      "model_slug": "office-skills-no-renderer",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "office-skills-no-renderer"
      ]
    },
    "parallel-tool-calls": {
      "model_slug": "parallel-tool-calls",
      "cosmos_test_count": 71,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "parallel-tool-calls"
      ]
    },
    "packed-chunks": {
      "model_slug": "packed-chunks",
      "cosmos_test_count": 63,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "packed-chunks"
      ]
    },
    "partial-landlock-child-failure": {
      "model_slug": "partial-landlock-child-failure",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "partial-landlock-child-failure"
      ]
    },
    "persistent-pwsh-padded-completion": {
      "model_slug": "persistent-pwsh-padded-completion",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "persistent-pwsh-padded-completion"
      ]
    },
    "persistent-pwsh-tool-turn": {
      "model_slug": "persistent-pwsh-tool-turn",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "persistent-pwsh-tool-turn"
      ]
    },
    "pi-ai-replay-metadata": {
      "model_slug": "pi-ai-replay-metadata",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "pi-ai-replay-metadata"
      ]
    },
    "plugin-manager": {
      "model_slug": "plugin-manager",
      "cosmos_test_count": 54,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "plugin-manager"
      ]
    },
    "plugin-manager-mcp": {
      "model_slug": "plugin-manager-mcp",
      "cosmos_test_count": 59,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "plugin-manager-mcp"
      ]
    },
    "product-subagent-both": {
      "model_slug": "product-subagent-both",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "product-subagent-both"
      ]
    },
    "product-subagent-codex": {
      "model_slug": "product-subagent-codex",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "product-subagent-codex"
      ]
    },
    "product-subagent-result-diagnostic": {
      "model_slug": "product-subagent-result-diagnostic",
      "cosmos_test_count": 105,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "product-subagent-result-diagnostic"
      ]
    },
    "provider-cwd": {
      "model_slug": "provider-cwd",
      "cosmos_test_count": 17,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "provider-cwd"
      ]
    },
    "ptc-node-read-only": {
      "model_slug": "ptc-node-read-only",
      "cosmos_test_count": 24,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-node-read-only"
      ]
    },
    "ptc-node-workspace": {
      "model_slug": "ptc-node-workspace",
      "cosmos_test_count": 24,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-node-workspace"
      ]
    },
    "ptc-python-turn": {
      "model_slug": "ptc-python-turn",
      "cosmos_test_count": 51,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-python-turn"
      ]
    },
    "ptc-read-image": {
      "model_slug": "ptc-read-image",
      "cosmos_test_count": 57,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-read-image"
      ]
    },
    "ptc-read-image-attachment-path": {
      "model_slug": "ptc-read-image-attachment-path",
      "cosmos_test_count": 73,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-read-image-attachment-path"
      ]
    },
    "ptc-workspace-context": {
      "model_slug": "ptc-workspace-context",
      "cosmos_test_count": 55,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-workspace-context"
      ]
    },
    "pty-tools-sandbox-backend": {
      "model_slug": "pty-tools-sandbox-backend",
      "cosmos_test_count": 83,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "pty-tools-sandbox-backend"
      ]
    },
    "pwsh-tool-turn": {
      "model_slug": "pwsh-tool-turn",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "pwsh-tool-turn"
      ]
    },
    "ralph-loop": {
      "model_slug": "ralph-loop",
      "cosmos_test_count": 123,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ralph-loop"
      ]
    },
    "read-image": {
      "model_slug": "read-image",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image"
      ]
    },
    "read-image-attachment-path": {
      "model_slug": "read-image-attachment-path",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image-attachment-path"
      ]
    },
    "read-image-dimension": {
      "model_slug": "read-image-dimension",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image-dimension"
      ]
    },
    "read-image-reencode": {
      "model_slug": "read-image-reencode",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image-reencode"
      ]
    },
    "read-image-gif": {
      "model_slug": "read-image-gif",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image-gif"
      ]
    },
    "read-image-text-route": {
      "model_slug": "read-image-text-route",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "read-image-text-route"
      ]
    },
    "repeat-tool-reminder": {
      "model_slug": "repeat-tool-reminder",
      "cosmos_test_count": 87,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "repeat-tool-reminder"
      ]
    },
    "session-query-spill": {
      "model_slug": "session-query-spill",
      "cosmos_test_count": 80,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "session-query-spill"
      ]
    },
    "session-reference-spill": {
      "model_slug": "session-reference-spill",
      "cosmos_test_count": 35,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "session-reference-spill"
      ]
    },
    "skill-load": {
      "model_slug": "skill-load",
      "cosmos_test_count": 68,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "skill-load"
      ]
    },
    "session-sandbox-root": {
      "model_slug": "session-sandbox-root",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "session-sandbox-root"
      ]
    },
    "subagent-child-question-rejection": {
      "model_slug": "subagent-child-question-rejection",
      "cosmos_test_count": 89,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-child-question-rejection"
      ]
    },
    "subagent-acp-diagnostic": {
      "model_slug": "subagent-acp-diagnostic",
      "cosmos_test_count": 57,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-acp-diagnostic"
      ]
    },
    "subagent-depth-two-rejection": {
      "model_slug": "subagent-depth-two-rejection",
      "cosmos_test_count": 135,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-depth-two-rejection"
      ]
    },
    "subagent-max-tokens-partial": {
      "model_slug": "subagent-max-tokens-partial",
      "cosmos_test_count": 91,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-max-tokens-partial"
      ]
    },
    "subagent-multi": {
      "model_slug": "subagent-multi",
      "cosmos_test_count": 166,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-multi"
      ]
    },
    "subagent-published-run-failure": {
      "model_slug": "subagent-published-run-failure",
      "cosmos_test_count": 52,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-published-run-failure"
      ]
    },
    "subagent-parallel": {
      "model_slug": "subagent-parallel",
      "cosmos_test_count": 119,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-parallel"
      ]
    },
    "subagent-tool-filter": {
      "model_slug": "subagent-tool-filter",
      "cosmos_test_count": 46,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "subagent-tool-filter"
      ]
    },
    "team-targets": {
      "model_slug": "team-targets",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "team-targets"
      ]
    },
    "tool-call-turn": {
      "model_slug": "tool-call-turn",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "tool-call-turn"
      ]
    },
    "todo-write": {
      "model_slug": "todo-write",
      "cosmos_test_count": 45,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "todo-write"
      ]
    },
    "web-search-endpoint-guidance": {
      "model_slug": "web-search-endpoint-guidance",
      "cosmos_test_count": 45,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "web-search-endpoint-guidance"
      ]
    },
    "web-fetch": {
      "model_slug": "web-fetch",
      "cosmos_test_count": 43,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "web-fetch"
      ]
    },
    "windows-acl-skill": {
      "model_slug": "windows-acl-skill",
      "cosmos_test_count": 23,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "windows-acl-skill"
      ]
    },
    "workflow-confinement": {
      "model_slug": "workflow-confinement",
      "cosmos_test_count": 24,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "workflow-confinement"
      ]
    },
    "workflow-run": {
      "model_slug": "workflow-run",
      "cosmos_test_count": 87,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "workflow-run"
      ]
    },
    "workspace-dependencies": {
      "model_slug": "workspace-dependencies",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "workspace-dependencies"
      ]
    },
    "workspace-edit": {
      "model_slug": "workspace-edit",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "workspace-edit"
      ]
    },
    "approval-composer": {
      "model_slug": "approval-composer",
      "cosmos_test_count": 98,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "approval-composer"
      ]
    },
    "auto-review-denial": {
      "model_slug": "auto-review-denial",
      "cosmos_test_count": 20,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "auto-review-denial"
      ]
    },
    "auto-review-approval": {
      "model_slug": "auto-review-approval",
      "cosmos_test_count": 20,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "auto-review-approval"
      ]
    },
    "changed-files-turn": {
      "model_slug": "changed-files-turn",
      "cosmos_test_count": 84,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "changed-files-turn"
      ]
    },
    "code-language": {
      "model_slug": "code-language",
      "cosmos_test_count": 26,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "code-language"
      ]
    },
    "cordis-tool-round": {
      "model_slug": "cordis-tool-round",
      "cosmos_test_count": 105,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cordis-tool-round"
      ]
    },
    "deepseek-messages-chat": {
      "model_slug": "deepseek-messages-chat",
      "cosmos_test_count": 50,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "deepseek-messages-chat"
      ]
    },
    "diff-bounded": {
      "model_slug": "diff-bounded",
      "cosmos_test_count": 27,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "diff-bounded"
      ]
    },
    "feedback-command": {
      "model_slug": "feedback-command",
      "cosmos_test_count": 59,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "feedback-command"
      ]
    },
    "file-upload-round": {
      "model_slug": "file-upload-round",
      "cosmos_test_count": 68,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "file-upload-round"
      ]
    },
    "fresh-round-trip": {
      "model_slug": "fresh-round-trip",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "fresh-round-trip"
      ]
    },
    "goal-multi-turn-actions": {
      "model_slug": "goal-multi-turn-actions",
      "cosmos_test_count": 251,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "goal-multi-turn-actions"
      ]
    },
    "lifecycle-chrome": {
      "model_slug": "lifecycle-chrome",
      "cosmos_test_count": 50,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "lifecycle-chrome"
      ]
    },
    "live-interactions": {
      "model_slug": "live-interactions",
      "cosmos_test_count": 50,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "live-interactions"
      ]
    },
    "markdown-file-links": {
      "model_slug": "markdown-file-links",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "markdown-file-links"
      ]
    },
    "message-feedback-protocol": {
      "model_slug": "message-feedback-protocol",
      "cosmos_test_count": 14,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "message-feedback-protocol"
      ]
    },
    "minimal-preset": {
      "model_slug": "minimal-preset",
      "cosmos_test_count": 62,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "minimal-preset"
      ]
    },
    "navigation-panes": {
      "model_slug": "navigation-panes",
      "cosmos_test_count": 49,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "navigation-panes"
      ]
    },
    "permission-policy-context": {
      "model_slug": "permission-policy-context",
      "cosmos_test_count": 256,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "permission-policy-context"
      ]
    },
    "plan-review": {
      "model_slug": "plan-review",
      "cosmos_test_count": 82,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "plan-review"
      ]
    },
    "present": {
      "model_slug": "present",
      "cosmos_test_count": 86,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "present"
      ]
    },
    "present-svg": {
      "model_slug": "present-svg",
      "cosmos_test_count": 58,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "present-svg"
      ]
    },
    "preset-migration": {
      "model_slug": "preset-migration",
      "cosmos_test_count": 11,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "preset-migration"
      ]
    },
    "ptc-escalation-approved": {
      "model_slug": "ptc-escalation-approved",
      "cosmos_test_count": 66,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-escalation-approved"
      ]
    },
    "pwsh-terminal": {
      "model_slug": "pwsh-terminal",
      "cosmos_test_count": 23,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "pwsh-terminal"
      ]
    },
    "ptc-round": {
      "model_slug": "ptc-round",
      "cosmos_test_count": 77,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ptc-round"
      ]
    },
    "question-composer": {
      "model_slug": "question-composer",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "question-composer"
      ]
    },
    "schedule-catalog": {
      "model_slug": "schedule-catalog",
      "cosmos_test_count": 22,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "schedule-catalog"
      ]
    },
    "seeded-history": {
      "model_slug": "seeded-history",
      "cosmos_test_count": 33,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "seeded-history"
      ]
    },
    "steering": {
      "model_slug": "steering",
      "cosmos_test_count": 80,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "steering"
      ]
    },
    "tool-details": {
      "model_slug": "tool-details",
      "cosmos_test_count": 100,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "tool-details"
      ]
    },
    "turn-tail-actions": {
      "model_slug": "turn-tail-actions",
      "cosmos_test_count": 65,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "turn-tail-actions"
      ]
    },
    "web-search-round": {
      "model_slug": "web-search-round",
      "cosmos_test_count": 71,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "web-search-round"
      ]
    },
    "session": {
      "model_slug": "session",
      "cosmos_test_count": 101,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "SESSION",
        "session"
      ]
    },
    "sessions-experts": {
      "model_slug": "sessions-experts",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "SESSIONS_EXPERTS"
      ]
    },
    "s": {
      "model_slug": "s",
      "cosmos_test_count": 494,
      "pass_rate_pct": 5.1,
      "twelve_cs_score": 9.3,
      "passes": 25,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "s"
      ]
    },
    "c": {
      "model_slug": "c",
      "cosmos_test_count": 180,
      "pass_rate_pct": 6.1,
      "twelve_cs_score": 10.3,
      "passes": 11,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "c"
      ]
    },
    "quota": {
      "model_slug": "quota",
      "cosmos_test_count": 7,
      "pass_rate_pct": 85.7,
      "twelve_cs_score": 85.9,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "quota"
      ]
    },
    "h": {
      "model_slug": "h",
      "cosmos_test_count": 37,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "h"
      ]
    },
    "spend": {
      "model_slug": "spend",
      "cosmos_test_count": 16,
      "pass_rate_pct": 37.5,
      "twelve_cs_score": 40.1,
      "passes": 6,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "spend"
      ]
    },
    "meter": {
      "model_slug": "meter",
      "cosmos_test_count": 3,
      "pass_rate_pct": 100.0,
      "twelve_cs_score": 99.5,
      "passes": 3,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "meter"
      ]
    },
    "p": {
      "model_slug": "p",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "p"
      ]
    },
    "projection": {
      "model_slug": "projection",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "projection"
      ]
    },
    "m": {
      "model_slug": "m",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "m"
      ]
    },
    "live-ccr": {
      "model_slug": "live-ccr",
      "cosmos_test_count": 641,
      "pass_rate_pct": 0.2,
      "twelve_cs_score": 4.7,
      "passes": 1,
      "fails": 170,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "live_ccr"
      ]
    },
    "08-nemo35f": {
      "model_slug": "08-nemo35f",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "08_nemo35f"
      ]
    },
    "10-lagunaxs": {
      "model_slug": "10-lagunaxs",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 7,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "10_lagunaxs"
      ]
    },
    "14-lingvl": {
      "model_slug": "14-lingvl",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "14_lingvl"
      ]
    },
    "18-gemma426b": {
      "model_slug": "18-gemma426b",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 7,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "18_gemma426b"
      ]
    },
    "19-qwen27f": {
      "model_slug": "19-qwen27f",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 5,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "19_qwen27f"
      ]
    },
    "24-inklingf": {
      "model_slug": "24-inklingf",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "24_inklingf"
      ]
    },
    "26-gemma431b": {
      "model_slug": "26-gemma431b",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 7,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "26_gemma431b"
      ]
    },
    "ledger": {
      "model_slug": "ledger",
      "cosmos_test_count": 750,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "ledger"
      ]
    },
    "control": {
      "model_slug": "control",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "control"
      ]
    },
    "cosmos-cc-1p7yx0v": {
      "model_slug": "cosmos-cc-1p7yx0v",
      "cosmos_test_count": 204,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cosmos_cc__1p7yx0v"
      ]
    },
    "q": {
      "model_slug": "q",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "q"
      ]
    },
    "cosmos-claude-rail-e1wk6d5g": {
      "model_slug": "cosmos-claude-rail-e1wk6d5g",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cosmos_claude_rail_e1wk6d5g"
      ]
    },
    "claude-rail": {
      "model_slug": "claude-rail",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "claude_rail"
      ]
    },
    "collector": {
      "model_slug": "collector",
      "cosmos_test_count": 58,
      "pass_rate_pct": 58.6,
      "twelve_cs_score": 60.2,
      "passes": 34,
      "fails": 2,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "collector"
      ]
    },
    "queue": {
      "model_slug": "queue",
      "cosmos_test_count": 14,
      "pass_rate_pct": 14.3,
      "twelve_cs_score": 18.1,
      "passes": 2,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "queue"
      ]
    },
    "export": {
      "model_slug": "export",
      "cosmos_test_count": 17,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "export"
      ]
    },
    "cosmos-code-export": {
      "model_slug": "cosmos-code-export",
      "cosmos_test_count": 28,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "cosmos_code_export"
      ]
    },
    "hero-run-20261008": {
      "model_slug": "hero-run-20261008",
      "cosmos_test_count": 2154,
      "pass_rate_pct": 36.5,
      "twelve_cs_score": 39.2,
      "passes": 786,
      "fails": 186,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {},
      "source_clusters": [
        "_hero_run_20261008"
      ]
    },
    "apodex-apodex-1.1-mini-free": {
      "model_slug": "apodex-apodex-1.1-mini-free",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 4,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "apodex_apodex-1.1-mini_free"
      ]
    },
    "door-opencode": {
      "model_slug": "door-opencode",
      "cosmos_test_count": 53,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3
      },
      "source_clusters": [
        "door_opencode"
      ]
    },
    "nvidia-nemotron-3-nano-omni-30b-a3b-reasoning-free": {
      "model_slug": "nvidia-nemotron-3-nano-omni-30b-a3b-reasoning-free",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "EMPTY": 3,
        "LOOKS_SHAPED": 2
      },
      "source_clusters": [
        "nvidia_nemotron-3-nano-omni-30b-a3b-reasoning_free"
      ]
    },
    "qwen-qwen3.5-9b": {
      "model_slug": "qwen-qwen3.5-9b",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 7,
        "EMPTY": 2,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "qwen_qwen3.5-9b"
      ]
    },
    "ibm-granite-granite-4.0-h-micro": {
      "model_slug": "ibm-granite-granite-4.0-h-micro",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 4
      },
      "source_clusters": [
        "ibm-granite_granite-4.0-h-micro"
      ]
    },
    "google-gemma-4-26b-a4b-it-free": {
      "model_slug": "google-gemma-4-26b-a4b-it-free",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 8,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 8
      },
      "source_clusters": [
        "google_gemma-4-26b-a4b-it_free"
      ]
    },
    "amazon-nova-lite-v1": {
      "model_slug": "amazon-nova-lite-v1",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FENCE": 10
      },
      "source_clusters": [
        "amazon_nova-lite-v1"
      ]
    },
    "inference-net-schematron-v2-small": {
      "model_slug": "inference-net-schematron-v2-small",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 7,
        "FENCE": 2,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "inference-net_schematron-v2-small"
      ]
    },
    "google-lyria-3-pro-preview": {
      "model_slug": "google-lyria-3-pro-preview",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 9,
        "UPSTREAM": 1
      },
      "source_clusters": [
        "google_lyria-3-pro-preview"
      ]
    },
    "google-gemma-3-4b-it": {
      "model_slug": "google-gemma-3-4b-it",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FENCE": 2,
        "LOOKS_SHAPED": 2
      },
      "source_clusters": [
        "google_gemma-3-4b-it"
      ]
    },
    "gryphe-mythomax-l2-13b": {
      "model_slug": "gryphe-mythomax-l2-13b",
      "cosmos_test_count": 6,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 6
      },
      "source_clusters": [
        "gryphe_mythomax-l2-13b"
      ]
    },
    "qwen-qwen3-32b": {
      "model_slug": "qwen-qwen3-32b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 2,
        "LOOKS_SHAPED": 2
      },
      "source_clusters": [
        "qwen_qwen3-32b"
      ]
    },
    "google-gemma-4-31b-it-free": {
      "model_slug": "google-gemma-4-31b-it-free",
      "cosmos_test_count": 8,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 8,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 8
      },
      "source_clusters": [
        "google_gemma-4-31b-it_free"
      ]
    },
    "inference-net-schematron-v2-turbo": {
      "model_slug": "inference-net-schematron-v2-turbo",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 10
      },
      "source_clusters": [
        "inference-net_schematron-v2-turbo"
      ]
    },
    "google-lyria-3-clip-preview": {
      "model_slug": "google-lyria-3-clip-preview",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 9
      },
      "source_clusters": [
        "google_lyria-3-clip-preview"
      ]
    },
    "cohere-north-mini-code-free": {
      "model_slug": "cohere-north-mini-code-free",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 2,
        "LOOKS_SHAPED": 2
      },
      "source_clusters": [
        "cohere_north-mini-code_free"
      ]
    },
    "inclusionai-ling-3.0-flash-fin": {
      "model_slug": "inclusionai-ling-3.0-flash-fin",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 3,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "inclusionai_ling-3.0-flash-fin"
      ]
    },
    "ibm-granite-granite-4.2-8b": {
      "model_slug": "ibm-granite-granite-4.2-8b",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 2,
        "MOUTH_FORM": 2
      },
      "source_clusters": [
        "ibm-granite_granite-4.2-8b"
      ]
    },
    "inclusionai-ling-3.0-flash-sante": {
      "model_slug": "inclusionai-ling-3.0-flash-sante",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 2,
        "LOOKS_SHAPED": 2
      },
      "source_clusters": [
        "inclusionai_ling-3.0-flash-sante"
      ]
    },
    "poolside-laguna-xs-2.1-free": {
      "model_slug": "poolside-laguna-xs-2.1-free",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 2,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 2,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "poolside_laguna-xs-2.1_free"
      ]
    },
    "nvidia-nemotron-3.5-content-safety-free": {
      "model_slug": "nvidia-nemotron-3.5-content-safety-free",
      "cosmos_test_count": 9,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "NOT_CODE": 7,
        "MOUTH_FORM": 2
      },
      "source_clusters": [
        "nvidia_nemotron-3.5-content-safety_free"
      ]
    },
    "openai-gpt-5-nano": {
      "model_slug": "openai-gpt-5-nano",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 2,
        "MOUTH_FORM": 2
      },
      "source_clusters": [
        "openai_gpt-5-nano"
      ]
    },
    "microsoft-phi-4": {
      "model_slug": "microsoft-phi-4",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "FENCE": 10
      },
      "source_clusters": [
        "microsoft_phi-4"
      ]
    },
    "meta-muse-spark-1.2-contributor": {
      "model_slug": "meta-muse-spark-1.2-contributor",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 2,
        "REJECTED": 1,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "meta_muse-spark-1.2-contributor"
      ]
    },
    "rekaai-reka-edge": {
      "model_slug": "rekaai-reka-edge",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 8,
        "FENCE": 1,
        "MOUTH_FORM": 1
      },
      "source_clusters": [
        "rekaai_reka-edge"
      ]
    },
    "rekaai-reka-flash-3": {
      "model_slug": "rekaai-reka-flash-3",
      "cosmos_test_count": 19,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 16,
        "EMPTY": 3
      },
      "source_clusters": [
        "rekaai_reka-flash-3"
      ]
    },
    "meta-muse-spark-1.3-contributor": {
      "model_slug": "meta-muse-spark-1.3-contributor",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 2,
        "REJECTED": 1,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "meta_muse-spark-1.3-contributor"
      ]
    },
    "nvidia-nemotron-3.5-lightning": {
      "model_slug": "nvidia-nemotron-3.5-lightning",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 2,
        "MOUTH_FORM": 2
      },
      "source_clusters": [
        "nvidia_nemotron-3.5-lightning"
      ]
    },
    "meta-llama-llama-3.2-1b-instruct": {
      "model_slug": "meta-llama-llama-3.2-1b-instruct",
      "cosmos_test_count": 10,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 10
      },
      "source_clusters": [
        "meta-llama_llama-3.2-1b-instruct"
      ]
    },
    "google-gemma-3-27b-it": {
      "model_slug": "google-gemma-3-27b-it",
      "cosmos_test_count": 3,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 1,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 2,
        "UPSTREAM_POOL": 1
      },
      "source_clusters": [
        "google_gemma-3-27b-it"
      ]
    },
    "morph-morph-v3-large": {
      "model_slug": "morph-morph-v3-large",
      "cosmos_test_count": 5,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 4,
        "REJECTED": 1
      },
      "source_clusters": [
        "morph_morph-v3-large"
      ]
    },
    "google-gemma-2-27b-it": {
      "model_slug": "google-gemma-2-27b-it",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "LOOKS_SHAPED": 4
      },
      "source_clusters": [
        "google_gemma-2-27b-it"
      ]
    },
    "minimax-minimax-m1": {
      "model_slug": "minimax-minimax-m1",
      "cosmos_test_count": 26,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 23,
        "LOOKS_SHAPED": 2,
        "EMPTY": 1
      },
      "source_clusters": [
        "minimax_minimax-m1"
      ]
    },
    "mancer-weaver": {
      "model_slug": "mancer-weaver",
      "cosmos_test_count": 18,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 17,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "mancer_weaver"
      ]
    },
    "meta-llama-llama-guard-4-12b": {
      "model_slug": "meta-llama-llama-guard-4-12b",
      "cosmos_test_count": 15,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 0,
      "median_latency_ms": 145.0,
      "top_scars": {
        "MOUTH_FORM": 12,
        "NOT_CODE": 3
      },
      "source_clusters": [
        "meta-llama_llama-guard-4-12b"
      ]
    },
    "mistralai-mistral-saba": {
      "model_slug": "mistralai-mistral-saba",
      "cosmos_test_count": 4,
      "pass_rate_pct": 0.0,
      "twelve_cs_score": 0.0,
      "passes": 0,
      "fails": 0,
      "rate_limits_429": 3,
      "median_latency_ms": 145.0,
      "top_scars": {
        "UPSTREAM_POOL": 3,
        "LOOKS_SHAPED": 1
      },
      "source_clusters": [
        "mistralai_mistral-saba"
      ]
    }
  }
}