# CodeVetter

> Execution-backed verification and evaluation for coding agents. CodeVetter turns an agent change into reproducible runtime evidence and a measurable verdict without uploading the repository to a CodeVetter server.

## Product

- [Home](https://codevetter.com/): Features, pricing, how it works
- [Public benchmark](https://codevetter.com/benchmark): Reproducible cases, scorer, results, and limitations
- [Documentation](https://codevetter.com/docs/): Product, architecture, development, and operations truth
- [Download](https://codevetter.com/download): macOS release
- [FAQ](https://codevetter.com/faq): Common questions about local verification and provider boundaries
- [Changelog](https://codevetter.com/changelog): Verified shipped outcomes
- [Privacy](https://codevetter.com/privacy): Local-first data handling

## Verification guides

- [Coding-agent verification](https://codevetter.com/coding-agent-verification): Category guide to task, change, execution, evidence, and verdict
- [How to verify AI-generated code](https://codevetter.com/verify-ai-generated-code): Practical task-to-evidence workflow
- [AI code review vs verification](https://codevetter.com/ai-code-review-vs-verification): Comparison of static review and executable verification
- [Verification evidence bundle](https://codevetter.com/verification-evidence-bundle): Reference for portable verification artifacts

## Machine surfaces

- [Agent catalog](https://codevetter.com/api/ai): JSON inventory of public surfaces
- [Homepage markdown](https://codevetter.com/index.md): Product brief without JS
- [Benchmark markdown](https://codevetter.com/benchmark.md): Reproducible benchmark summary without JS
- [Documentation markdown](https://codevetter.com/docs.md): Canonical documentation index without JS
- [This index](https://codevetter.com/llms.txt)
- [Full agent brief](https://codevetter.com/llms-full.txt)