[
    {
      "schema_version": "amplio-failure-checkpoint.v1",
      "id": "checkpoint-001-controller-timeout",
      "status": "failed",
      "title": "FAILED \u2014 controller timeout; main agent stopped at step 7",
      "run_id": "2195hynj",
      "session_id": "main-agent",
      "step": 7,
      "occurred_at": "2026-09-20T09:25:33.461Z",
      "recorded_at": "2026-09-20T09:29:37.610415+00:00",
      "error": "ControllerLifecycleError: inference controller exited with request_timeout while the worker was active.",
      "scope": "Terminal infrastructure failure before any submitted decision, accepted action or settled day.",
      "recovery_note": "The supervisor cancelled the agent through the Amplio API. The raw event says \"cancelled via UI\", but this was controller cleanup, not a user cancellation. Call 15 returned HTTP 200 and native usage; all 15 calls are settled. A provider outage is not proved.",
      "source_event": {
        "session_id": "main-agent",
        "step": 7,
        "occurred_at": "2026-09-20T09:25:33.461Z",
        "event": {
          "content": "cancelled via UI",
          "marker": "cancelled",
          "type": "system"
        }
      },
      "controller_failure": {
        "error_type": "ControllerLifecycleError",
        "message": "Inference controller ended without a completed worker: request_timeout",
        "occurred_at": "2026-09-20T09:26:04.446292+00:00",
        "phase": "supervise",
        "status": "error",
        "traceback": "Traceback (most recent call last):\n  File \"/Users/eddy/dev/amplio/evals/structured-credit-trader/back-to-back-funding-book/src/worker_boundary.py\", line 806, in _run_worker\n    _require(not finished.is_set(), \"Inference controller stopped while the worker was active\")\n    ~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/Users/eddy/dev/amplio/evals/structured-credit-trader/back-to-back-funding-book/src/worker_boundary.py\", line 65, in _require\n    raise BoundaryError(message)\nworker_boundary.BoundaryError: Inference controller stopped while the worker was active\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n  File \"/Users/eddy/dev/amplio/evals/structured-credit-trader/back-to-back-funding-book/tools/run_worker_smoke.py\", line 669, in run_smoke\n    result = supervise(FinancialGate(runtime, gate, supervisor), supervisor, [family_root],\n                       policy=POLICY, max_attempts=MAX_ATTEMPTS, worker_timeout=WORKER_TIMEOUT,\n                       exposure_manifest=exposure_manifest,\n                       backend=\"bedrock\", region=REGION)\n  File \"/Users/eddy/dev/amplio/evals/structured-credit-trader/back-to-back-funding-book/src/worker_boundary.py\", line 985, in supervise\n    worker = _run_worker(record_path, manifest, attempt[\"artifact_name\"], policy, worker_timeout,\n                         attempt_dir / \"worker.json\", api_key_env, backend=backend, region=region)\n  File \"/Users/eddy/dev/amplio/evals/structured-credit-trader/back-to-back-funding-book/src/worker_boundary.py\", line 834, in _run_worker\n    raise ControllerLifecycleError(\"Inference controller ended without a completed worker: \" + reason)\nworker_boundary.ControllerLifecycleError: Inference controller ended without a completed worker: request_timeout\n"
      },
      "controller_outcome": {
        "controller": {
          "backend": "bedrock",
          "mode": "host-mediated-inference",
          "models": [
            "global.openai.gpt-6-astra"
          ],
          "ok": true,
          "reason": "request_timeout",
          "region": "us-east-1",
          "requested_reasoning_efforts": {
            "max": 15
          },
          "stats": {
            "calls_started": 15,
            "delivered": 14,
            "expired_results": 0,
            "output_tokens_reported": 7104,
            "output_tokens_reserved": 614400,
            "upstream_failures": 1,
            "upstream_successes": 14
          }
        },
        "controller_error_type": null,
        "controller_exited_while_worker_active": true,
        "monitor_exit_code": 124,
        "run_id": "2195hynj",
        "status": "controller_error"
      },
      "accepted_actions": 0,
      "elapsed_days": 0,
      "provider_timeout_proved": false,
      "source_db_sha256": "7e79225e5a9c914cdfd4023931d88c465215d24f11278cae5b8e8aa24f8cb7d6"
    }
  ]