Investigate p99 latency regression

auth-gateway · claude-haiku-4-5 · 52m · 2d ago · Blake Vidal

Part of: Auth token refresh race

Session tokens

91K tokens

272K raw tokens

Avoidable tokens (est.)

11K tokens

12% of session

Quality

craft check score

Outcome

Completed

5 cycles · 1 corrections · 52m wall-clock

PromptGenerateVerifyCorrect

6maccepted

11maccepted

13mrejected

10maccepted

8maccepted

correction

5 verify events · 0 accepts without verify — chips are pinned to the cycles that triggered them.

Grounded answers with mandatory cycle citations

Answers are PROXY — generated over the measured transcript. This query is visible to the engineer in their access log.

What drove the score — pinned to its cycles

Verification discipline98
Every accepted diff was preceded by a verify-type event
Plan-mode discipline32
No plan-mode use in this session
Correction load74
1 of 5 cycles were corrections of earlier output
Prompt clarity63
Context, constraints, and a done-condition raise first-pass acceptance

Caught in this session

low
Policy violation
Attempted edit outside allowed workspace path
Flagged in-session by your risk rules
low
PII in transcript
Customer email addresses appeared in transcript
Flagged in-session by your risk rules