Guides

Reviewing code your agent wrote.

Methods first. arnx is one way to apply them.

Each guide answers one question, gives a workflow you can repeat, and says where the method stops working. The commands were checked against the current arnx release. The methods hold with any tool.

• AI code review: comments versus executable evidence

  What AI code review is, how comment-based and evidence-based review differ, where each fits, and how to judge a tool.

• How to review code written by Claude Code

  Claude Code's own review commands, what they do not run, and how to add tests that execute through the arnx plugin or skill.

• How to review code written by Codex

  Reviewing Codex's changes with tests that run: where Codex reads the arnx skill, what we have and have not tested, and a manual flow for any agent.

• AI code review before you commit or open a pull request

  Reviewing before the commit or pull request: what gets reviewed, why not a Git hook, and the cheap check you can repeat in CI.

• How to verify AI-generated code

  A verification checklist for agent-written code and the four levels of evidence, from a failing assertion to a violation of intended behavior.

• Keeping regression tests from AI review findings

  Turning a confirmed finding into a retained regression test: what to read first, where the test lives, and how edits are detected.

Working through an agent? Start with arnx for coding agents. For evidence that the approach finds real defects, and how often it is wrong, read the case studies and the research notes.