Code review agent
Reviews every pull request for bugs, security issues and style, and comments on GitHub.
What it is
An AI agent you run on your own machine or server. You tell it the goal; it plans the steps, runs the commands and checks the result, asking before anything risky.
How it works
- Download the agent and its config.
- Give it access to the server or account it should work on.
- Describe the job; review and approve its plan.
- It does the work and leaves a log of every step.
What you get
- The agent, ready to run
- Works with your own Claude or OpenAI key
- Setup guide
- Download link and license by email, VAT handled by Lemon Squeezy
Free skills and prompts for the same job
- Code review — Reviews a diff for real bugs first, style last, and reports each finding with a concrete failure case.
- Pull request description — Turns a branch into a PR description reviewers can read in one minute: what, why, how to test, risks.
- Systematic debugging — Finds the root cause of a bug with hypotheses and evidence instead of random edits.
Learning the basics first? Read the KuanAI blog or try the free AI tools.
Coding agent skillsDrop-in skills for Claude Code and other coding agents: reviews, refactors, tests and deploys.$19SEO blog agentResearches keywords, writes and publishes articles to your blog on a schedule, with your review.$59QA testing agentClicks through your web app like a user, finds broken flows and writes the bug report.$59