Overview
Tool failure rate
≥ 2.9 %
1 of 35 results
Unreported tool outcomes are counted per session only, so the fleet figure is a lower bound. 1 calls never returned a result and are in neither figure.
Spend
≥ $1.90
16 responses carried no cost.
Worst stall
43.3 s
Blocked
≥ 2
Lower bound — 2 of 3 sessions predate this measurement and are not included.
What enforcement decided
- 2 Blocked refused before the call ran
- 1 Allowed no rule denied it
- 0 Redacted ran with arguments rewritten
- 0 Held for approval not an allow and not a block
- 0 Unclassified recorded, but carried no readable verdict
Worst repeat run ≥ 0 calls — back-to-back identical calls in one session. 2 sessions could not be read for this and are excluded.
3 sessions in this window. 1 excluded from the worst-stall and repeat figures because their events did not arrive in execution order. 1 excluded from the repeat figure because a tool call had no measurable identity. Both are normal for push sources.
Which tools fail most?
| Tool | Calls | Failure rate | Unanswered | Max duration |
|---|---|---|---|---|
| Read | 15 | 0 | 245 ms | |
| Bash | 17 | 1 | 689 ms | |
| Grep | 3 | 0 | 16 ms | |
| Glob | 1 | 0 | 210 ms |
Where is the spend going?
| Model | Requests | Cost | No cost reported | Input (uncached) | Cache reads | Cache writes | Output |
|---|---|---|---|---|---|---|---|
| claude-opus-5 | 31 | 16 | 94 | 1,673,204 | 108,654 | 28,338 |
What is going wrong?
No errors. 3 sessions checked.