- Outcome
- PASS3m
- Reward
- 100.0%
- Checks passed
- 25 / 25
- Runtime
- 3m 23s
- Harness
- Cursor CLI
- Task version
- v3.1.0
Checks
| Check | Result | Credit | Weight |
|---|---|---|---|
| baseline: the sync job publishes Slack documents | PASS | 100% | 1% |
| baseline: the MCP server reads a published Slack document | PASS | 100% | 1% |
| refetch: a new or edited top-level post rebuilds its month without fetching any replies | PASS | 100% | 9% |
| refetch: a new reply fetches only its own thread, once | PASS | 100% | 11% |
| refetch: deleted replies and a deleted parent rebuild from stored replies | PASS | 100% | 8% |
| refetch: a renamed channel rebuilds every month without fetching any replies | PASS | 100% | 8% |
| raw: every message the history listing returns is stored as Slack returned it | PASS | 100% | 7% |
| raw: every thread's replies are stored as Slack returned them | PASS | 100% | 7% |
| raw: after changes the store holds Slack's current version of what changed | PASS | 100% | 6% |
| retention: under retention, hidden months stay put and a visible reply fetches only its thread | PASS | 100% | 8% |
| retention: once retention lifts, a reply in a formerly hidden month fetches only its thread | PASS | 100% | 6% |
| interrupted: after a run fails partway, the next runs are complete and a later change fetches only its thread | PASS | 100% | 12% |
| deploy: over a mirror the base wrote, a month's next change fetches only its thread | PASS | 100% | 11% |
| page size: history is listed at 999 per page and replies at the largest page | PASS | 100% | 5% |
| preserve: Slack documents | PASS | 100% | 0% |
| preserve: Slack placement | PASS | 100% | 0% |
| preserve: incremental changes | PASS | 100% | 0% |
| preserve: retention | PASS | 100% | 0% |
| preserve: interrupted run | PASS | 100% | 0% |
| preserve: deploy over an existing mirror | PASS | 100% | 0% |
| preserve: channel removal | PASS | 100% | 0% |
| preserve: MCP search | PASS | 100% | 0% |
| preserve: MCP read_document | PASS | 100% | 0% |
| preserve: the base's own backend tests | PASS | 100% | 0% |
| preserve: backend ruff and mypy --strict | PASS | 100% | 0% |