{
  "run_id": "ShivSupane--fliprate-claude-haiku-5-5-1793108006656786519",
  "checker": {
    "claims": [
      {
        "claim": "Clean messy text directly with the normalize command",
        "command": "<guest-path> normalize 'mala product avadlaaa nhi'",
        "evidence": "mala product avadla nahi\n  avadlaaa -> avadla\n  nhi -> nahi\n",
        "expected": "mala product avadla nahi",
        "input": "mala product avadlaaa nhi",
        "kind": "claim",
        "outcome": "Real input produced the expected result: mala product avadla nahi",
        "returncode": 0,
        "source": "fliprate normalize \"mala product avadlaaa nhi\"",
        "source_checked": true,
        "source_file": "README.md",
        "verdict": "yes"
      },
      {
        "claim": "Gate decision TRANSFORM means unstable but fixed by the built-in normalizer",
        "command": "<guest-path> gate 'mala product avadlaaa nahi'",
        "evidence": "TRANSFORM: Stable after approved normalization\nprediction='product_feedback' raw_stability=0.625 normalized_stability=1.0\n  flipped on: mala produkt avdla nahi\n  flipped on: mala prodct avdla nahee\n  flipped on: mala produuuct avadlaaa nahi\n",
        "expected": "TRANSFORM: Stable after approved normalization",
        "input": "mala product avadlaaa nahi",
        "kind": "claim",
        "outcome": "Real input produced the expected result: TRANSFORM: Stable after approved normalization",
        "returncode": 0,
        "source": "| **TRANSFORM** | Unstable, but the built-in normalizer fixes it.",
        "source_checked": true,
        "source_file": "README.md",
        "verdict": "yes"
      }
    ],
    "cycle_checks": [
      {
        "claim": "Scheduled or looping main behavior",
        "command": "",
        "evidence": "Inspected README (CLI, API, production deployment sections), CI workflow, pyproject entry points and src/fliprate. No cron, scheduler, --once/--now, or webhook trigger exists. The only while True loop (src/fliprate/events.py drain) is a generator consumer, not a product cycle. FastAPI jobs are request-driven.",
        "expected": "identify the product trigger",
        "input": "source inspection",
        "kind": "cycle",
        "outcome": "Inspected README (CLI, API, production deployment sections), CI workflow, pyproject entry points and src/fliprate. No cron, scheduler, --once/--now, or webhook trigger exists. The only while True loop (src/fliprate/events.py drain) is a generator consumer, not a product cycle. FastAPI jobs are request-driven.",
        "source": "repository entry points",
        "verdict": "yes"
      }
    ],
    "details": {
      "argv": [
        "<guest-path>",
        "normalize",
        "mala product avadlaaa nhi"
      ],
      "kind": "cli",
      "needs_browser": false,
      "needs_credentials": false,
      "returncode": 0,
      "scope": "product operation",
      "stderr": "",
      "stdout": "mala product avadla nahi\n  avadlaaa -> avadla\n  nhi -> nahi",
      "working": true
    },
    "evidence_strength": "readme_checked",
    "example_checks": [
      {
        "claim": "Basic usage example runs the library quick-check, normalization and audit flow",
        "command": "<guest-path> <guest-path>",
        "deferred": false,
        "evidence": "on: Prediction is stable across safe variants\n\nMessy: 'Mala product avadlaaa nhiii!!!'\nClean: 'Mala product avadla nahi!'\n\nSafe prediction on 'ye delivery nhi aayi': negative\n\nAuditing model on dataset...\nMessages: 4 | safe variants tested: 12\nAccuracy on originals:      75.0%  (macro-F1 0.78)\nAccuracy on safe variants:  75.0%  (macro-F1 0.78)\nAccuracy drop:              +0.0%  95% CI [+0.0%, +0.0%]\nFlip rate:                  0.0%  95% CI [0.0%, 0.0%]\nMost fragile change type:   spelling (flip rate 0.0%)\nNote: flip rate shows instability, not correctness. A model can flip to the right answer.",
        "expected": "prints a clean text, safe prediction and audit report",
        "input": "examples/01_basic_usage.py with no arguments",
        "kind": "example",
        "outcome": "command completed successfully",
        "returncode": 0,
        "source": "examples/01_basic_usage.py (docs/QUICKSTART.md Python API)",
        "verdict": "yes"
      }
    ],
    "status": "PASS"
  },
  "diagnosis": null
}