Adds `fineenvs_v3/` only; nothing else in the repository changes. Please review the README text before merging.
- Evaluated the 999 FineEnvs v2 tasks without a successful attempt: Qwen/Qwen3.8-27B (thinking on), 2 attempts per task, Terminus-2, 32,768-token context, 8,192 output tokens per call.
- 515 newly kept tasks (`tasks.parquet`, disjoint from the 4,001 v2 tasks; 4,516 combined); 484 still without success (`archive.parquet`). Pass@1 44.89%, pass@2 51.55%; success counts 0/2: 484, 1/2: 133, 2/2: 382.
- `traces/`: all 1,998 attempts (successful and failed) with conversations and result records; 6 attempts without a verifier score (task memory limit), counted as unsuccessful.
- `trace_stats/`, `plots/`, `metadata/` (selection, results, settings audit of all 1,998 saved configurations, contract, stage reports, release verification) and `patch.py`, which reproduces both task files from the pinned v2 archive.
- Task payloads are unchanged from the evaluated v2 archive rows (content ha…
Publish PyMethods2Test v1: 4,990 kept tasks with verified images
Verify all 27 rerun payload hashes against the 14,564 already-published tasks. Update current discovery status and README coverage; preserve underlying run evidence and unchanged task payloads.
Restore recorded Tmax validation timings and publish timing evidence
Restore three infrastructure build cases and seven reviewed no-op cases. Publish the v1 README and preserve acceptance decisions and original diagnostic evidence in metadata.
Record user-approved AI-detection exclusion for Tmax v1