- Outcome
- PASS22m
- Reward
- 100.0%
- Checks passed
- 22 / 22
- Runtime
- 21m 43s
- Harness
- Cursor CLI
- Task version
- v3.1.2
Checks
| Check | Result | Credit | Weight |
|---|---|---|---|
| baseline: the app serves and a prompt set's responses are searchable | PASS | 100% | 1% |
| baseline: the reports list answers for a workspace | PASS | 100% | 1% |
| batch returns each analyzed tag's single-tag analysis and none for the other tags | PASS | 100% | 11% |
| batch selects each tag's answers exactly as a run's tag filter does | PASS | 100% | 4% |
| batch only looks analyses up and never starts one | PASS | 100% | 4% |
| batch finds a tag's single-tag analysis when the tag has more responses than the sample cap | PASS | 100% | 11% |
| batch honours extra prompt sets in the scope | PASS | 100% | 4% |
| batch honours the run window | PASS | 100% | 4% |
| batch honours the search query | PASS | 100% | 4% |
| a run's citation map lists each domain's cited URLs with the responses that cite them | PASS | 100% | 11% |
| cited URLs are stored and come back when the analysis is read by id or from the cache | PASS | 100% | 8% |
| batch items carry their analysis's cited URLs | PASS | 100% | 4% |
| cited URLs list exactly the analyzed responses when a run is down-sampled | PASS | 100% | 4% |
| the reports list counts narratives across every tag | PASS | 100% | 8% |
| a published per-tag report's page shows each tag and its analysis | PASS | 100% | 8% |
| the analysis page shows a selected tag's analysis from the batch | PASS | 100% | 4% |
| the analysis page runs the analysis for a selected tag | PASS | 100% | 4% |
| the analysis page publishes a report only once every tag has an analysis | PASS | 100% | 8% |
| preserve: a narratives run, its cache and its cache-only lookup | PASS | 100% | 0% |
| preserve: an analysis is read and deleted by id | PASS | 100% | 0% |
| preserve: reports are published, listed, renamed and deleted | PASS | 100% | 0% |
| preserve: the app's pages | PASS | 100% | 0% |