* fix: coerce stringified numeric anchor in timeline() to repair MCP anchor routing
HTTP query params arrive as strings, so the typeof anchor === 'number'
dispatch always missed the observation-ID branch, falling through to
ISO-timestamp parsing and silently returning a wrong-epoch window with
the correct anchor echoed in the header. Closes the timeline regression
reported on cut-guardian.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* refactor: extract parseNumericAnchor helper and expand timeline tests
Address CodeRabbit review nitpicks on PR #2176:
- Extract anchor coercion into private parseNumericAnchor helper
- Add whitespace-padded numeric-string anchor test case
- Add explicit numeric-anchor-not-found regression test
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* test: assert exact ordering and rendered anchor header in timeline tests
Address CodeRabbit nitpick on PR #2176: drop sort to verify chronological
ordering, and assert that the rendered anchor/header text echoes the
requested numeric ID and marks the anchor row.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* test: extract anchor-render helper and tighten garbage-anchor assertion
Address CodeRabbit nitpicks: DRY-up the three repeated anchor header/row
assertions into expectAnchorRendered(), and assert the exact
"Invalid timestamp: 123abc" error in the garbage-anchor branch instead
of a generic non-empty-string check.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Alex Newman