mapping-codebases
Scannednpx machina-cli add skill oaustegard/claude-skills/mapping-codebases --openclawMapping Codebases
Generate _MAP.md files providing hierarchical code structure views. Maps show function signatures, class methods, imports, and line numbers—enabling API understanding without reading full source files.
Installation
uv venv /home/claude/.venv
uv pip install tree-sitter-language-pack --python /home/claude/.venv/bin/python
Generate Maps
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo --skip tests,.github,locale
Common skip patterns: tests,.github,.husky,locale,migrations,__snapshots__,coverage,target,docs
Navigate Via Maps
After generating maps, use them for navigation—read _MAP.md files, not source files directly.
Workflow:
- Read root
_MAP.mdfor high-level structure - Follow subdirectory links to drill into relevant areas
- Use function signatures and class methods to understand APIs
- Read full source only when implementation details are needed
Maps reveal without reading source:
- All public functions with signatures:
def record_attempt(subtask_id, session, success, approach, error=None):200 - Class structure with methods:
RecoveryManager→classify_failure(),is_circular_fix(),rollback_to_commit() - Import relationships showing module dependencies
- Line numbers for direct navigation
Anti-pattern: Reading files directory-by-directory. Use maps to find what you need, then read only the specific file/lines required.
Map Contents
Each _MAP.md includes:
- Directory statistics (file/subdirectory counts)
- Subdirectory links for hierarchical navigation
- Per-file: imports preview, symbol hierarchy with (C)lass/(m)ethod/(f)unction markers
- Function signatures (Python, TypeScript, Go, Rust, Ruby)
- Line numbers (
:42format) for every symbol - Markdown files: h1/h2 heading ToC
- Other files section (JSON, YAML, configs)
Example:
# services/
*Files: 4 | Subdirectories: 0*
## Files
### recovery.py
> Imports: `json, subprocess, dataclasses, datetime, enum`...
- **FailureType** (C) :24
- **RecoveryManager** (C) :43
- **__init__** (m) `(self, spec_dir: Path, project_dir: Path)` :55
- **classify_failure** (m) `(self, error: str, subtask_id: str)` :137
- **record_attempt** (m) `(subtask_id, session, success, approach, error=None)` :200
- **is_circular_fix** (m) `(self, subtask_id: str, current_approach: str)` :242
- **get_recovery_hints** (m) `(self, subtask_id: str)` :495
Commands
# Generate maps
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo
# Skip directories
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo --skip tests,.github
# Clean maps
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo --clean
# Dry run
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo -n
# Verbose (debug output)
/home/claude/.venv/bin/python /mnt/skills/user/mapping-codebases/scripts/codemap.py /path/to/repo -v
Default skips: .git, node_modules, __pycache__, .venv, venv, dist, build, .next
Supported Languages
Python, JavaScript, TypeScript, TSX, Go, Rust, Ruby, Java, HTML, Markdown.
Limitations
- Structural info only (symbols/imports), not semantic descriptions
- Signatures: Python (full), TypeScript/Go/Rust/Ruby (params + return types), Java (not extracted)
- Markdown: h1/h2 headings only
- Private symbols (
_prefix) excluded from top-level exports
Source
git clone https://github.com/oaustegard/claude-skills/blob/main/mapping-codebases/SKILL.mdView on GitHub Overview
Generates _MAP.md files per directory that list exports, imports, classes, and methods along with signatures and line numbers. This helps you quickly understand unfamiliar repositories and APIs without reading every source file.
How This Skill Works
The skill uses tree-sitter to parse source code, extracting symbols, imports, and line numbers. It then generates per-directory _MAP.md files with a hierarchical view, marking Class (C), Function (f), and Method (m) symbols and including signature previews for quick navigation.
When to Use It
- When exploring a repository to understand its structure and API surface
- Before making modifications in an unfamiliar codebase
- To analyze module dependencies via import relationships
- When starting work on an unfamiliar repository to map its layout
- To navigate efficiently after generation by reading _MAP.md rather than raw source files
Quick Start
- Step 1: Create a Python virtual environment and install the tree-sitter language pack
- Step 2: Run the codemap script on your repo, e.g. /path/to/repo --skip tests,.github,locale
- Step 3: Open root _MAP.md and use the links to drill into relevant directories
Best Practices
- Run the mapping process on a clean clone and re-run after large changes
- Use the default skip patterns or customize with --skip to focus on relevant areas
- Start at the root _MAP.md and drill into subdirectories for targeted exploration
- Verify symbol signatures against actual code when precision matters
- Use the imports previews to understand module dependencies before edits
Example Use Cases
- Map a Python web app to view public functions and class methods with signatures and line numbers
- Map a Go microservice to inspect exported packages, funcs, and types
- Map a TypeScript/TSX project to understand React components and utilities via imports
- Map a Rust project to see module structure, crates, and function signatures
- Map a Java project to quickly grasp API surfaces across packages