## FAIL

### Blocking findings

1. **Technical mappings are still called “confirmed.”**  
   `scripts/health/dashboard_v5/read_api.py` sets day provenance `review_status` to `"confirmed"` whenever mapping coverage is complete. `dashboard-v5.js:867` renders that as `Review: bestätigt`, even though no explicit user confirmation is required.  
   **Action:** distinguish technical completeness from user verification, e.g. `technically_complete` / “technisch vollständig – nicht bestätigt”; reserve “bestätigt” for explicit reviewed mappings.

2. **Sprint 7C-D browser spec is not in the regular isolated browser matrix.**  
   `tests/browser/run_v5_isolated_matrix.sh` contains neither a `7c-d` group nor `dashboard_v5_sprint7c_d.spec.js`. Adding it only to `package.json` does not put it into the repository’s regular synthetic matrix.  
   **Action:** add a proper 7C-D fixture/setup group and include it in the default matrix sequence.

3. **Browser acceptance currently fails: 3 passed, 3 failed.**  
   In the synthetic browser run, the action form was absent, so technical-provenance, preview, and defer/queue scenarios could not execute. The responsive/compact, filter, and Today-task scenarios passed.  
   **Action:** configure the 7C-D matrix fixture with mapping actions enabled, then require all six scenarios to pass.

4. **Preview assertion disagrees with the intended separated presentation.**  
   Runtime correctly emits separate lines `Kategorie: …` and `SIGHi: …`, but the browser spec expects the combined literal `Kategorie SIGHi`. Once the action form is enabled, that assertion will fail.  
   **Action:** assert `Kategorie:` and `SIGHi:` independently.

### Criteria that look correct

- Suggestions and technical assignments are otherwise labeled separately from explicit confirmation.
- Preview code separately shows canonical mapping, actual catalog category, SIGHi value, source, entry count, and day count.
- Examples/days use complete counts with bounded lists and explicit truncation disclosure; focused unit coverage passes.
- No new stress semantics, derived stress score, or causal interpretation was introduced. The ambiguity and stop decision are documented, and no Stress/Anspannung Today task appears.
- Compact viewport/200%-text scenario passed, but action-control usability remains unproven because the action form was missing.

### Verification

- Focused Python tests: **21 passed**.
- Syntax/diff checks: Python compilation, JavaScript syntax, and `git diff --check` passed.
- Browser: **3 passed / 3 failed**.
- Read-only review used synthetic data only; **no repository files modified**. Temporary Playwright artifacts were removed.