Your code is under new management. Agents that review your code - locally or on every PR - using the Skills you already know and love.
Skills, not prompts. Define analysis once, run it anywhere. Bootstrap your environment with skills from conventional directories (.agents/skills/ or .claude/skills/).
Two ways to run. CLI catches issues before you push. GitHub Action reviews every PR automatically.
GitHub-native. Findings appear as inline PR comments with suggested fixes.
# Initialize warden in your repository
npx warden init
# Run on uncommitted changes
# Uses Claude Code subscription if logged in, or set WARDEN_ANTHROPIC_API_KEY
npx warden
# Fix issues automatically
npx warden --fixgit clone [email protected]:getsentry/warden.git
cd warden
pnpm install && pnpm build
pnpm test # unit tests
pnpm test:evals # end-to-end evals (requires API key)See evals/README.md for the eval framework.
FSL-1.1-ALv2