Repository navigation

[#985][feat][eval] Add per-phase time tracking and planning timeout - #987

Open

ayazhankadessova wants to merge 12 commits into

mainfrom

Open

[#985][feat][eval] Add per-phase time tracking and planning timeout#987
ayazhankadessova wants to merge 12 commits into
mainfrom
issue-985

Commits on Apr 6, 2026
- [docs] Document planning timeout, per-phase timing, and eval folder README
  Ayazhan Kadessova
  committed
  97034b1
- [test] Add per-phase timing and planning_timeout tests
  Ayazhan Kadessova
  committed
  17bc175
- [feat][eval] Add per-phase time tracking and --planning-timeout flag
  Ayazhan Kadessova
  committed
  fb2e1fa
Commits on Apr 7, 2026
- [feat][eval] Plumb phase timing from daemon thread via timing_bucket
  Ayazhan Kadessova
  committed
  399044f
- [feat][eval] Enforce --planning-timeout in full mode via sub-thread
  Ayazhan Kadessova
  committed
  af694ef
- [feat][eval] Persist per-task results to results.jsonl
  Ayazhan Kadessova
  committed
  5f72743
- [docs][eval] Add phase timing breakdown from 5-task sample run
  Ayazhan Kadessova
  committed
  44c2f08
- [docs][eval] Add per-mode phase timing sections with placeholders for nlcmd and codex
  Ayazhan Kadessova
  committed
  d42cae0
- [docs][eval] Fill nlcmd and codex impl phase timing from partial runs
  Ayazhan Kadessova
  committed
  cec0399
- [docs][eval] Add raw opus row to resolve rates table
  Ayazhan Kadessova
  committed
  16fbb5d
- [docs][eval] Update raw opus resolve rate to 11/20 (55%) from SWE-bench scoring
  Ayazhan Kadessova
  committed
  cd4511d
Commits on Apr 14, 2026
- [docs][eval] Update codex impl 20-task results: 8/20 completed, 12 Codex API crashes
  Ayazhan Kadessova
  committed
  40c0448