| Tool | Use When |
|---|---|
Serena find_symbol |
Know the symbol name - TRY FIRST |
Serena find_referencing_symbols |
Find all usages of a symbol |
Serena get_symbols_overview |
Understand file structure |
rg "pattern" |
Regex/text patterns (not symbol-based) |
Built-in Grep / Glob |
Fallback when above tools insufficient |
| Tool | Use When |
|---|---|
Serena replace_symbol_body |
Replacing entire methods/functions |
Serena insert_after_symbol |
Adding new code after a symbol |
Built-in Edit |
All other edits |
Serena first, then rg/built-in tools.
4-stage evaluation framework testing Claude Code plugin component triggering (skills, agents, commands, hooks, MCP servers).
Requirements: Node.js >= 20.0.0, ANTHROPIC_API_KEY in .env
npm run build && npm run lint && npm run typecheck && npm run format:check && npm run knip && npm test # Verify
cc-plugin-eval run -p ./plugin # Full pipeline
cc-plugin-eval run -p ./plugin --dry-run # Cost estimation
cc-plugin-eval resume -r <run-id> # Resume run| Stage | Purpose | Entry |
|---|---|---|
| 1. Analysis | Parse plugin, extract triggers | src/stages/1-analysis/index.ts |
| 2. Generation | Create test scenarios (LLM + deterministic) | src/stages/2-generation/index.ts |
| 3. Execution | Run via Claude Agent SDK with tool capture | src/stages/3-execution/index.ts |
| 4. Evaluation | Programmatic detection → LLM judge fallback | src/stages/4-evaluation/index.ts |
SDKs: @anthropic-ai/sdk (Stages 2, 4), @anthropic-ai/claude-agent-sdk (Stage 3)
- Task guides:
docs/*.md- API usage, adding components, CI/CD, hooks/MCP notes - Serena memories:
.serena/memories/- architecture, testing, code style
- Detection: Programmatic (100%) → LLM judge (quality assessment fallback)
- Sessions: Default
batched_by_component(scenarios sharing a component reuse session with/clear) - State migration: Update
migrateState()insrc/state/operations.tsfor new component types