Triage Analysis Skill
Codebase Exploration Strategy
When analyzing a ticket against the codebase, follow this exploration order:
- •Understand the project: Read README, package.json/pyproject.toml/Cargo.toml (or equivalent) to understand the tech stack and structure.
- •Map the architecture: Identify entry points, main modules, and the dependency graph. Check for
references/memory.mdin the plugin directory for pre-defined repo relationships and domain terms. - •Search for relevance: Use grep/glob to find code related to the ticket's keywords, feature area, or error messages.
- •Trace the call chain: From relevant entry points, trace how data flows through the affected area.
- •Check test coverage: Look for existing tests in the affected area to understand current coverage and expected behavior.
- •Identify ripple effects: Check what imports/depends on the affected code to assess blast radius.
Impact Assessment Matrix
| Factor | Weight | How to Assess |
|---|---|---|
| Files changed | High | Count of files likely needing modification |
| Test coverage | High | Are affected areas well-tested? |
| Dependency depth | Medium | How deep in the dependency chain? |
| API surface | High | Does this change public APIs? |
| Data model | High | Does this change data schemas? |
| Cross-repo | High | Does this affect other repos? (check memory) |
Output Expectations
The triage analysis should produce:
- •A map of affected files with confidence levels
- •Architecture context (patterns in use, frameworks, conventions)
- •Complexity estimate: Simple (<1 day) / Moderate (1-3 days) / Complex (3+ days)
- •Risk factors specific to this codebase
Memory Integration
Always check references/memory.md for:
- •Domain glossary: Terms that have specific meanings in this organization
- •Repo map: How repositories relate to each other (shared libraries, microservices, monorepo structure)
- •Team conventions: Naming patterns, branching strategies, review requirements
- •Known debt: Areas of technical debt that affect estimation