[codex] add failure diagnosis contract

[divo12]

Summary

• add an immutable FailureDiagnosis tied to an authoritative failed OutcomeEvaluation
• define five stable failure types and supported/inconclusive evidence invariants
• bound identifiers, diagnosis text, and cited Langfuse observations
• export the provider-agnostic contract from the public package

Non-goals

• no failure miner or judge
• no Langfuse failure writer
• no report generation, pattern clustering, or taxonomy registry

Local verification

• uv sync --frozen --extra dev --extra plugin
• uv run ruff check src tests plugins/openflywheel/scripts/mcp_server.py
• uv run mypy src tests plugins/openflywheel/scripts/mcp_server.py
• uv run pytest --cov=ofw --cov-report=term-missing --cov-fail-under=90 -q: 150 passed, 2 skipped, 95.30% coverage
• radon: new contract maximum A (5), repository average A (2.36)
• trace-query-planner and outcome-recorder skill validation
• plugin validation

All verification listed above is local evidence, not CI evidence.

---

Summary by cubic

Adds a provider-agnostic FailureDiagnosis contract for failure mining, tied to an authoritative failed OutcomeEvaluation, and exports it from the ofw package.

Contract

• Defines five stable failure types and supported/inconclusive evidence invariants.
• Bounds identifiers, diagnosis text, and cited Langfuse observations.
• Raises a typed FailureDiagnosisError with a stable error code on invalid input.
• Leaves failure mining, judging, and report generation out of scope.

^{Written for commit ad68880. Summary will update on new commits.}

[cubic-dev-ai]

All reported issues were addressed across 5 files

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>