codeowners-coverage 0.2.0

Measure and enforce CODEOWNERS coverage

CODEOWNERS Coverage

A tool to measure and enforce CODEOWNERS coverage in your repository.

Prerequisites

• Python 3.9+
• Git repository
• Ollama (for suggest command) - Install from ollama.ai

Ollama Setup (Optional - for AI-powered suggestions)

The suggest command uses a local LLM (Ollama) to intelligently match files to teams based on git history.

1. Install Ollama: https://ollama.ai/
2. Pull the default model:

   ollama pull llama3.2
   

3. Ensure Ollama is running:

   ollama serve  # Runs on http://localhost:11434
   

4. Verify setup:

   just ollama-check
   # or manually:
   curl http://localhost:11434
   

Quick setup:

just ollama-setup  # Downloads and configures Ollama model

Installation

pip install codeowners-coverage

Usage

Check coverage

codeowners-coverage check

Generate baseline

codeowners-coverage baseline

Suggest CODEOWNERS entries (AI-powered)

Use local LLM to intelligently suggest team ownership based on git history.

GitHub Token Setup:

1. Create a Personal Access Token at https://github.com/settings/tokens
2. Choose either:
   • Classic Token: Select read:org scope
   • Fine-grained Token: Grant "Read-only" access to "Organization permissions > Members"
3. Export the token:

   export GITHUB_TOKEN=ghp_your_token_here
   

Usage:

# Basic usage (requires GITHUB_TOKEN env var)
# By default, includes both new uncovered files AND baseline files
export GITHUB_TOKEN=ghp_xxxxx
codeowners-coverage suggest

# Or with justfile
just suggest

# Only suggest for new uncovered files (exclude baseline)
codeowners-coverage suggest --no-baseline

# Auto-apply suggestions (creates backup)
codeowners-coverage suggest --apply

# Export as diff for review
codeowners-coverage suggest --format=diff > suggestions.patch

# Configure Ollama model
codeowners-coverage suggest --ollama-model llama3.2

# Adjust directory consolidation threshold
codeowners-coverage suggest --min-coverage 0.9  # 90% consistency required

How it works:

1. Finds all uncovered files (both new and baseline by default)
2. Analyzes git history to find contributors per file
3. Fetches GitHub teams and maps contributors to teams
4. Uses local LLM to intelligently suggest ownership
5. Consolidates to high-level directory patterns when appropriate

Requirements:

• Ollama running locally (setup: just ollama-setup)
• GitHub Personal Access Token (set via GITHUB_TOKEN env var)
  • Token type: Classic PAT or Fine-grained PAT
  • Required scope: read:org (to read organization teams and members)
  • Token holder must be a member of the organization
  • Create at: https://github.com/settings/tokens
• Git repository with commit history

Note: By default, suggestions include both new uncovered files AND baseline files. Use --no-baseline to only suggest for new uncovered files.

Configuration

Create a .codeowners-config.yml file in your repository root:

# Path to CODEOWNERS file
codeowners_path: ".github/CODEOWNERS"

# Path to baseline file
baseline_path: ".github/codeowners-coverage-baseline.txt"

# File patterns to exclude from coverage checking
exclusions:
  - "**/__pycache__/**"
  - "**/*.pyc"
  - "node_modules/**"
  - "dist/**"
  - ".venv/**"

# Suggestion settings (for 'suggest' command)
github_token: ${GITHUB_TOKEN}  # GitHub Personal Access Token (needs read:org scope)
github_org: mycompany  # Auto-detected from git remote if omitted
ollama_model: llama3.2  # Ollama model to use
ollama_base_url: http://localhost:11434  # Ollama API endpoint
suggestion_min_coverage: 0.8  # Min % for directory consolidation
suggestion_lookback_commits: 100  # How far back to analyze git history

GitHub Actions Integration

Add a workflow to check CODEOWNERS coverage on pull requests:

name: CODEOWNERS Coverage
on: [pull_request]

jobs:
  check:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      - run: pip install codeowners-coverage
      - run: codeowners-coverage check

Features

• Pattern Matching: Supports all CODEOWNERS patterns (wildcards, globstars, etc.)
• Baseline Support: Track progress incrementally with a baseline allowlist
• AI-Powered Suggestions: Use local LLM (Ollama) to intelligently suggest CODEOWNERS entries based on git history and team membership
• Smart Consolidation: Automatically groups file-level suggestions into high-level directory patterns
• Configurable: Customize exclusions and paths via YAML config
• Fast: Uses pathspec library for efficient pattern matching
• GitHub Actions Ready: Easy integration with CI/CD pipelines

Development

Using just (recommended)

This project includes a justfile with common development commands:

# Install package in development mode
just install

# Run all tests
just test

# Run tests with coverage report
just test-cov

# Run type checking
just typecheck

# Run linting
just lint

# Run all checks (tests, typecheck, lint)
just check

# Clean build artifacts
just clean

# Build the package
just build

# Or build with uv (faster)
just build-uv

# Format code
just format

# Check Ollama setup
just ollama-check

# Setup Ollama (downloads model)
just ollama-setup

# Run suggest command
just suggest

# Publish to Test PyPI
just publish-test

# Publish to production PyPI
just publish

# Full release workflow
just release

Manual commands

# Install in development mode
uv pip install -e ".[dev]"

# Run tests
pytest tests/ -v

# Run with coverage
pytest tests/ -v --cov

# Build package
python -m build

# Publish to PyPI
twine upload dist/*

License

Apache-2.0