Show HN: Verity – self-healing review gate for Claude Code
Verity is a self-healing review gate for Claude Code that repairs code on every run, ensuring only clean and secure code moves forward. It features built-in memory to make each agent run smarter and stabilize costs. The tool is free in beta and installable via npm.
Background
- Claude Code is Anthropic's CLI-based agent that can autonomously edit and run code from a terminal. As more teams let AI agents directly modify codebases, the risk of introducing subtle bugs, security holes, or unwanted changes grows.
- "Adversarial review" means a second AI agent (or a rule-based system) actively tries to find problems in the code the first agent wrote — like a sparring partner that checks the work.
- "Self-healing" means the tool doesn't just flag issues; it automatically fixes them before the code is approved, so developers only see passing reviews.
- Verity is built by Codacy, a well-known code quality and security analysis platform used in many development workflows. This is their entry into the AI-agent tooling space.
- The package is installed via npm and integrates directly into Claude Code's workflow as a "gate" — meaning Claude Code can't proceed until Verity's checks pass, catching problems mid-run rather than after the fact.