| name | score-session |
| description | Score the current AI coding session for orchestration quality across 5 metrics |
Score Session
Analyze the current conversation and score it for AI orchestration quality.
Trigger
Use when you want to evaluate how well this coding session used AI orchestration patterns — tool diversity, decision-making, error recovery, and compound learning.
Workflow
- Gather the current session context — tools invoked, decisions made, errors encountered, skills used
- Use the
score_session MCP tool to score the session content
- Present the score breakdown with actionable feedback
Input
The full session transcript or a summary of the current conversation. If the user provides a specific session export file, use score_session_file instead.
Output
- Overall score (0-100) with quality tier (Exceptional/Excellent/Good/Fair/Poor)
- Per-metric breakdown:
- Skill Diversity (20%) — range of tools and skills invoked
- Decision Depth (25%) — explicit tradeoffs and reasoning
- Error Recovery (20%) — how errors were handled
- Compound Learning (20%) — cross-step insights
- Orchestration Mastery (15%) — agent coordination patterns
- Actionable suggestions for improving the score
Constraints
- Score reflects orchestration quality, not code correctness
- Sessions with no tool use or decisions will score low — that's expected for simple tasks
- Scores are deterministic for the same input