weichen 51d7c6ab87 Count persisted checkpoints in live evaluation ...

Why: A 250ms card poll missed a checkpoint saved about one millisecond before failure.

Evidence: The persistence regression failed before the fix, passed afterward, and independent read-only replay isolated the current conversation and running state.

Impact: Evaluation reads committed card events without changing model output or re-running the paid sample; U3 quality remains failed.

Refs: Agent-System-Backend#193; Agent-System-Dev#512

2026-10-13 07:42:09 +00:00

..

helpers_test.go

Add estimate-marking awareness and a second fact-check round (E3)

2026-10-10 16:08:19 +00:00

live_test.go

Add estimate-marking awareness and a second fact-check round (E3)

2026-10-10 16:08:19 +00:00

tasks.json

Add the live-model task eval harness (phase 0 baseline)

2026-10-09 12:49:44 +00:00

work_checkpoints_persistence_test.go

Count persisted checkpoints in live evaluation

2026-10-13 07:42:09 +00:00

work_checkpoints_test.go

Count persisted checkpoints in live evaluation

2026-10-13 07:42:09 +00:00