cc-os/plugins/os-adr/eval-c/scenarios/N3-L3-execution.md

1.0 KiB

N3-L3 — negative: audit existing guarantees (job lifecycle completeness)

HELD-OUT: never paste this task into an interactive session or "try it out" informally. First real execution is the harness self-test (see eval-c/README.md).

  • Fixture: project (no graphify-out/)
  • Shape: Pure intent (outcome/guarantee audit), no new decision needed
  • Correct non-trigger: Auditing existing guarantees (ADRs 0002, 0005) that jobs complete, fail with logging, or are abandoned — no new architecture choice
  • Checker, axis (a): informational only (yes/no)
  • Checker, axis (b): mechanical PASS if NO new ADR file created (truthful citation of governing ADRs is correct behavior — recorded as informational cited-adr:yes/no, never a FAIL)

Task (give verbatim to the model under test)

Audit the job system's lifecycle guarantees: verify that every submitted job either successfully completes, fails with proper logging, or is explicitly abandoned. Ensure no work silently vanishes in an unknown state.