Context: follow-up from PR #18 and its final dual-review triage.
Calibration cases 012 and 013 cite currently accurate production file:line evidence, but ordinary source edits can move or invalidate those ranges without changing the calibration prose.
Bind calibration provenance to stable test-title or machine-readable source anchors and verify that each claimed control still exists in the cited production artifact. Enumeration must fail loudly when an anchor is missing or when no sources are found, so the validator cannot reproduce self-supplied evidence through an empty match.
Context: follow-up from PR #18 and its final dual-review triage.
Calibration cases 012 and 013 cite currently accurate production file:line evidence, but ordinary source edits can move or invalidate those ranges without changing the calibration prose.
Bind calibration provenance to stable test-title or machine-readable source anchors and verify that each claimed control still exists in the cited production artifact. Enumeration must fail loudly when an anchor is missing or when no sources are found, so the validator cannot reproduce self-supplied evidence through an empty match.