Thanks for considering a contribution. Arkon is the AI governance platform — built for small teams and agency owners who run real agents in production.
Open an issue with:
- What you expected to happen
- What actually happened
- Steps to reproduce
- Your environment (OS, Docker version, browser)
Open an issue with the feature-request label. Explain the problem you're solving, not just the feature you want. "I need a way to stop an agent mid-run from my phone" is better than "add a mobile kill switch."
- Fork the repo
- Create a branch:
git checkout -b fix/your-fix-name - Make your changes
- Test locally with
docker compose up - Commit with a clear message:
fix: prevent duplicate threat alerts within cooldown window - Open a PR against
main
Use conventional commits:
fix:— bug fixesfeat:— new featuresdocs:— documentation changesrefactor:— code changes that don't fix bugs or add featureschore:— maintenance, dependencies, tooling
- TypeScript for all new code
- Follow existing patterns in the codebase
- Keep components focused — one responsibility per file
- API routes go in
src/app/api/ - Shared logic goes in
src/lib/
Check the issues for things tagged good-first-issue or help-wanted. High-impact areas:
- SDK development — Node.js and Python client libraries
- Threat detection patterns — new ThreatGuard regex patterns for emerging attack types
- Workflow templates — pre-built automation templates for common scenarios
- Documentation — guides, examples, translations
- NemoClaw/OpenClaw integration — deeper framework-specific features
Be direct, be helpful, be respectful. We're building tools for people who run AI agents in production — the stakes are real. Treat every contributor's time as valuable.
Open a discussion in the Q&A category or reach out on Discord.