why - Investigate the rationale behind code and design decisions
Investigates why code or design decisions exist by querying available evidence sources and returning cited findings with calibrated confidence.
Tags
Updated: 2026-10-08Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Discover available MCPs
- Anchor targets in code
- Search source control history
- Query evidence categories in parallel
- Synthesize cited findings
- Calibrate confidence levels
- Report evidence gaps
- Surface contradictory evidence
Inputs
- Investigation question
- Code target
- Repository files
- Git history
- Pull requests
- Available MCP servers
Outputs
- Cited investigation report
- Evidence coverage map
- Confidence-calibrated conclusions
- Reported null results
- Documented evidence gaps
- Surfaced alternative hypotheses
Requirements
- Git repository access
- GitHub CLI access
- Available evidence MCPs
- Cursor MCP environment
- Subagent model support
- Non-readonly agent mode