Skill Specification Quality Assessment on A2A-Agentization Bench (Stage 3)
70.4PrecisionClaude Code
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Claude Code2026.04 | 70.4 | 70.4 | 63 | |
| EnvX2026.04 | 69.8 | 74 | 66.2 | |
| OpenHands2026.04 | 65.2 | 69.7 | 59.9 | |
| Codex2026.04 | 50 | 47.8 | 42.1 |
| Method | Links | |||
|---|---|---|---|---|
| Claude Code2026.04 | 70.4 | 70.4 | 63 | |
| EnvX2026.04 | 69.8 | 74 | 66.2 | |
| OpenHands2026.04 | 65.2 | 69.7 | 59.9 | |
| Codex2026.04 | 50 | 47.8 | 42.1 |