Turn Success Rate
--%
--
-- complete / -- total
Session Completion
--%
--
-- completed / -- sessions
Tool Error Rate
--%
-- calls
Loop rate: --%
Workload Scope
12
JTBDs
Run: baseline
Evaluation Trends Over Time
Chronological trajectory of turn success, session completion, and tool errors
Execution Composition
Native response vs tool call distribution
Jobs-to-be-Done (JTBD) Workload Success Rates
Empirically derived from Reef production logs & Toolshed integration specs
| JTBD Category | Turns | Complete | Success Rate | 95% Wilson CI | Status |
|---|---|---|---|---|---|
| Loading JTBD data... | |||||
Live Production Turns Stream
Incoming agent turns streamed in real time from Reef Cloud SQL
Showing latest turns
| Time | Space | JTBD Classification | User Prompt Excerpt | Tools | Outcome |
|---|---|---|---|---|---|
| Loading incoming turns... | |||||
Historical Evaluation Runs
Loading runs...
Evaluation Job Queue
No active jobs.