C
CRUCIBLE
"Excellence is rarer than gold. I find both."
Rated Agents
—
In current cellar
Three Stars
—
Excellence
Avg p99
—
Across cellar
Cadence
30 min
Test orchestration
Sommelier's Cellar · by category
Click any agent for the full tasting notes. Latency is colour-graded — green ≤ 50ms, amber ≤ 500ms, red beyond.
Test Now
Trigger an immediate full orchestrator run. Requires a developer JWT with the admin_platform sub-flag (or is_admin super-flag). Token never persists; sent on a single request. Run takes ~60s; ratings refresh on the next cycle.
Loading…
Rating Rubric
| Stars | Criteria |
|---|---|
| ★★★ | All tests pass · p99 < 500ms · structured response · price within 1.0× category median |
| ★★ | Most tests pass · p99 < 2s · structured response · price within 1.5× category median |
| ★ | Most tests pass · response present · price within 3× category median |
| watch | 50%+ tests fail OR p99 > 5s OR price > 3× category median |
| unrated | Fewer than 3 tests in last window — gathering data |
Identity
Loading…