[PR #61] fix(mocks): repair end-to-end mock pipeline + numeric consistency

[il10241024]

🔗 Mirrored PR cyberfabric/cyber-insight-front#61 | Author: dzarlax | Opened: 2026-05-13T12:46:53Z | Status: open
GitHub API does not allow setting PR author or timestamps — attribution preserved here.

---

Reopens #99 (closed without merge on 2026-05-13; GitHub refuses API reopen). Same head branch, same commits, same diff.

Summary

Mock data pipeline had several silent failures that misled demos:

• Identity mocks bypassed: IdentityApiService.getPersonByEmail() calls baseURL-relative paths but identityMocks.ts only registered absolute URLs, so the mock was silently skipped.
• Mock handlers ignored OData filters: every team/IC drill returned the default fixture regardless of $filter, so navigating into different teams or members rendered the wrong synthetic data.
• Mock-bridge listener registered async: startup app/user/loaded could fire before the listener subscribed, leaving viewer/identity unset for leads & executives.
• Team headcounts were uniform (2/2/2/2/2/2) and exec-view header rendered raw i18n keys.
• Mock numbers contradicted each other — parent meeting_hours less than Teams + Zoom; merge_rate 88% with prs_merged 9 / prs_created 11; bugs_fixed 23 with bugs_to_task_ratio 18%; lines_per_commit 62 against clean_loc 12000 / commits 34; etc.
• Hours→days auto-scaling rendered a 60h meeting month as "3d", which loses the timesheet feel.
• Attention Needed list expanded to multiple rows per member when one person breached several thresholds — diverging from the "At Risk" KPI on the same screen.

Changes (4 commits)

1. fix(mocks): repair end-to-end mock pipeline broken on main — handler routing, filter dispatch, baseURL-relative identity-mock registration.
2. fix(mocks): vary team headcounts + ExecutiveView header fallback — 7 new PEOPLE entries (final per-team distribution 4/4/4/3/2/2), t('header.title' / 'header.subtitle') literal-key fallback, sub-lead chain so Bob Park's 6/6 stays canonical.
3. fix(mocks): address CodeRabbit review — three follow-ups from PR [PR #53] fix(mocks): repair end-to-end mock pipeline broken on main #104 review (identity baseURL relative paths, OData filter parsing, sync listener registration).
4. fix(mocks): make mock numbers internally consistent — see commit message for the full table; biggest items above.

Test plan

• tsc --noEmit clean
• eslint clean
• Architecture check passes (pre-commit)
• Local dev (VITE_ENABLE_MOCKS=1): exec-view → drill into each team → drill into each member; verify numbers shown match across views (e.g. meeting_hours = teams_meeting_hours + zoom_meeting_hours, prs_merged / prs_created ≈ merge_rate).
• Verify "At Risk" KPI count equals Attention Needed row count.
• Verify collaboration bullets stay in hours, not days.

Summary by CodeRabbit

Release Notes

• New Features

  • Added alert deduplication in team view—displays at most one alert per team member based on severity priority.

• Bug Fixes

  • Fixed team-size metrics calculation to accurately reflect current roster data.
  • Improved translation fallback handling in executive view header.

• Improvements

  • Collaboration metrics now retain hour units instead of auto-scaling to days.
  • Enhanced mock data with additional team members and updated metric configurations.

---