{
  "schema": "anvil-serving.agentic-result-summary/v1",
  "candidate": "thinkingcap-awq-5090",
  "evidence_root": "native/agentic",
  "scope": "bounded synthetic agentic diagnostic; not SWE-bench or real repository execution",
  "execution_topology": "co-resident C1",
  "results": {
    "synthetic_cases": 2,
    "passed_cases": 2,
    "api_turns": 6,
    "tool_calls": 4,
    "finish_reason_counts": {
      "tool_calls": 4,
      "stop": 2
    },
    "reasoning_disabled_observed_zero": true,
    "safe_error_retry": "passed"
  },
  "timing_interpretation": "No latency or software-engineering performance claim is made from this diagnostic.",
  "promoted": false,
  "redaction_ledger": "native/agentic/redactions.json"
}