List your company

Task Success on SkillLearnBench Random

29Success CountRevision v3
4.0410.521723.48May 31, 2026

Evaluation results

MethodLinks
Revision v3Model=GPT-5.5, Harness=Codex2026.052918
Revision v3Model=Opus-4.7, Harness=Claude Code2026.052540
Revision v3Model=DeepSeek-V4-Pro, Harness=Claude Code2026.052424
Skill v0Model=GPT-5.5, Harness=Codex2026.05236
Revision v3Model=Kimi-2.6, Harness=Claude Code2026.052232
No skillModel=GPT-5.5, Harness=Codex2026.0520—
Skill-CreatorModel=GPT-5.5, Harness=Codex2026.0517-6
Skill-CreatorModel=DeepSeek-V4-Pro, Harness=Claude Code2026.05168
Revision v3Model=Qwen-3.6plus, Harness=Claude Code2026.051520
Skill v0Model=DeepSeek-V4-Pro, Harness=Claude Code2026.05144
No skillModel=DeepSeek-V4-Pro, Harness=Claude Code2026.0512—
Skill v0Model=Kimi-2.6, Harness=Claude Code2026.051114.83
Skill-CreatorModel=Kimi-2.6, Harness=Claude Code2026.05108
Skill-CreatorModel=Qwen-3.6plus, Harness=Claude Code2026.0586
No skillModel=Opus-4.7, Harness=Claude Code2026.057—
Skill-CreatorModel=Opus-4.7, Harness=Claude Code2026.0570
Skill v0Model=Opus-4.7, Harness=Claude Code2026.0570
Skill v0Model=Qwen-3.6plus, Harness=Claude Code2026.0574
No skillModel=Kimi-2.6, Harness=Claude Code2026.056—
No skillModel=Qwen-3.6plus, Harness=Claude Code2026.055—