- Add renderProgressBar() helper in pi-extension.ts
- Pi init/reinit now show [████░░░░░░] 3/20 → filename.ts
- Replaces plain text percentage with visual bar
- All 52 tests passing
- Cache now lives at <project>/.cache/llm-cache.json (single hidden dir)
- Remove extra pi-project-map subdirectory level
- Use dirname() instead of string slicing for robustness
- Change cache location from .pi-project-map/cache/ to .cache/pi-project-map/
- Cache is now hidden by default (dot directory)
- Add .cache/ to .gitignore
- Remove committed .pi-project-map/cache/ from repo
- InitOptions.onProgress callback for directory-level progress
- generateDirectoryMap reports per-file progress (→ filename)
- initProject reports [N/M] Analyzing <path> (X files)...
- pi-extension wires _onUpdate to onProgress for live Pi UI updates
- User sees real-time feedback instead of silent long waits
- Use processFiles from llm-batch.ts in generateDirectoryMap
- Process files with 4 concurrent LLM calls + 100ms batch delays
- Reduces init time from O(n) sequential to O(n/4) parallel
- Prevents Pi tool timeout on projects with many files
- Call ctx.modelRegistry.getApiKeyAndHeaders(model) to resolve auth
- Pass resolved apiKey and headers to complete() options
- Provides clear error message if auth fails ( directs user to /login )
- Update test mock to include ok: true on auth response
- No external fetch() — Pi handles all transport and auth
- tree-sitter-go@0.23.4 peer depends on tree-sitter@^0.21.1
- Conflicts with tree-sitter@0.22.4 used by tree-sitter-python
- Go AST already falls back to generic export/deps extraction
- Add .npmrc with legacy-peer-deps=true as safety net
- npm install --omit=dev now works cleanly
- Upgrade tree-sitter to 0.22.4 with compatible grammar versions
- Add Python support via tree-sitter-python@0.23.6
- Add Go support via tree-sitter-go@0.23.4
- Extract class hierarchies with method signatures (params, returns)
- Extract call graphs including external library calls
- Extract exception raises (ValueError, Error, etc)
- Extract top-level functions with full signatures
- Encode rich AST data into compact markdown format:
class:Name, method:name(params)→return, call:..., raise:...
- Deduplicate nested call chains (db.query vs db.query().where())
- Filter out simple export names when rich func/class entries exist
- All 55 tests passing
- Cache now lives in <project>/.pi-project-map/cache/llm-cache.json
- Removed global ~/.cache/pi-project-map/ usage
- Cache travels with the project, no cross-project collisions
- Easy to invalidate: rm -rf .pi-project-map/cache/
Files changed:
- llm-cache.ts: accept cacheDir parameter, default to project-local
- llm-extract.ts: pass cacheDir through to cache functions
- init.ts: pass rootPath as cacheDir
- patch.ts: accept and pass cacheDir
- cli.ts: pass targetPath as cacheDir
- pi-extension.ts: pass ctx.cwd as cacheDir
- Tests updated to use temp dirs for cache isolation
- PiLLMClient now imports @mariozechner/pi-ai's complete() function
- Respects user's /model selection and /login auth
- No external fetch() — Pi handles transport internally
- Added src/types/pi-ai.d.ts for TypeScript declarations
- Extension reverted to real LLM calls (initProject, patchFile, reinitPath)
- Tests mock @mariozechner/pi-ai for verification
Inside Pi, the extension no longer attempts to call the LLM directly
(which Pi's ExtensionAPI doesn't support). Instead:
- project_map_init: generates maps using AST + heuristics (fast, free)
- project_map_patch: rewrites the file's directory with heuristics
- project_map_reinit: full heuristic regeneration
- project_map_validate: unchanged (no LLM needed)
The CLI still supports real LLM calls via --llm-provider=kimi|openai.
This separates concerns:
- Pi extension: deterministic structural analysis
- CLI: rich semantic analysis with configurable LLM
These grammars require tree-sitter@^0.25.0 but we use 0.21.1.
They already failed at runtime with try/catch graceful fallback.
Removing them fixes npm peer dependency conflict during pi install.
- Remove tree-sitter-go and tree-sitter-python from dependencies
- ast-extract.ts already handles missing grammars gracefully
- Heuristics + LLM layer cover Go/Python analysis anyway
- design-doc.md: Rewrote Section 3 (Pipeline Architecture) to describe the
real LLM integration: dual-provider LLM client (Pi native + external),
SHA-256 disk cache in ~/.cache/pi-project-map/, parallelization with
4-8 concurrent requests, retries with exponential backoff, hard-error
policy, context limit protection. Added new Section 4 with concrete
file-level and package-level prompts.
- implementation-plan.md: Replaced old milestone schedule with current
status and detailed M7 tasks for LLM integration: LLM client abstraction,
external API client, Pi LLM client, disk cache, parallel batching with
retries, rewriting llm-extract.ts, context limits, CLI/extension wiring,
and tests.
- Minor formatting cleanup in pi-extension.ts and src/cli.ts
All 16 tests pass. TypeScript compiles clean.
- format.ts: Full parsePackageMap implementation with round-trip support
- llm-extract.ts: Heuristic-based extraction without LLM API (regex exports,
imports, purpose inference from filename patterns)
- ast-extract.ts: Tree-sitter integration for TypeScript/TSX with fallback
for other languages
- init.ts: Refactored to expose generateDirectoryMap helper, reinitPath
delegates to initProject
- patch.ts: Small packages now use generateDirectoryMap for full rewrite
- discover.ts: Cleaned up require('fs') to use proper ES import
- Tests: 8 tests covering format round-trip and LLM heuristics
- All TypeScript compiles clean, all tests pass
- Design doc with full spec (dense markdown format, hybrid AST+LLM pipeline,
consumption model, stale data mitigation)
- Implementation plan with 6 milestones and rollout strategy
- TypeScript package structure with all source stubs
- CLI entry point, formatter, discover, init, patch, validate, extract, merge
- Pi SKILL.md with tool definitions and format documentation
alex