List your company

Environment setup and dependency repair on EnvBench Java (held-out)

38.32Task Pass RateCODESKILL
14.992821.048927.10533.1611May 25, 2026

Evaluation results

MethodLinks
CODESKILLSkill Backbone=Qwen3.5-4B, Frozen coding policy=Qwen3.5-35B-A3B2026.0538.325.3514
Prompt Skill Mgmt.Skill Backbone=GPT-5.4-mini, Frozen coding policy=Qwen3.5-35B-A3B2026.0536.455.4913.72
SFT-CODESKILLSkill Backbone=Qwen3.5-4B, Frozen coding policy=Qwen3.5-35B-A3B2026.0535.517.8513.65
Subtask MemorySkill Backbone=GPT-5.4-mini, Frozen coding policy=Qwen3.5-35B-A3B2026.0532.714.2513.34
Prompt Skill Mgmt.Skill Backbone=Qwen3.5-4B, Frozen coding policy=Qwen3.5-35B-A3B2026.0530.844.6713.09
No-skill baselineSkill Backbone=-, Frozen coding policy=Qwen3.5-35B-A3B2026.0527.16.114.03
CODESKILLSkill Backbone=Qwen3.5-4B, Frozen coding policy=GPT-5.4-mini2026.0527.11.799.1
Subtask MemorySkill Backbone=GPT-5.4-mini, Frozen coding policy=GPT-5.4-mini2026.0525.233.499.64
Prompt Skill Mgmt.Skill Backbone=GPT-5.4-mini, Frozen coding policy=GPT-5.4-mini2026.0524.32.279.35
SFT-CODESKILLSkill Backbone=Qwen3.5-4B, Frozen coding policy=GPT-5.4-mini2026.0523.242.218.25
Prompt Skill Mgmt.Skill Backbone=Qwen3.5-4B, Frozen coding policy=GPT-5.4-mini2026.0522.432.569.25
No-skill baselineSkill Backbone=-, Frozen coding policy=GPT-5.4-mini2026.0515.893.629.76