# Sphinx — AI Coding Tool Review & Pricing

> Sphinx is an open-source Git hook for agent-assisted development that pauses pushes to protected branches and quizzes you on what the change actually does. A separate reviewer model reads the commit through read-only Git tools, without seeing the agent conversation, and asks open-book questions about behavior, failure modes, and invariants before the push continues.

Source: https://ai.dosa.dev/tools/sphinx · Updated Oct 30, 2026 · Category: AI Code Review & Security · Published by dosa.dev

## At a glance

- **Tool**: Sphinx
- **Company**: parthh01
- **Category**: [AI Code Review & Security](https://ai.dosa.dev/category/code-review)
- **Pricing model**: Open Source
- **Official website**: https://sphinx-six-theta.vercel.app

## Key features

- Pre-push understanding check on protected branches (main/master by default)
- Independent reviewer model that never sees the coding agent's conversation
- Follow-up scenarios and explanations when an answer misses
- Works with Claude Code, Codex, Cursor, OpenCode, Pi, or plain Git
- Reviewer runs on Claude Code/Codex plans, OpenRouter, or any OpenAI-compatible endpoint

## Pricing

**Model**: Open Source

Free and open source under the MIT License; reviewer model calls run on your own Claude Code or Codex plan, OpenRouter key, or OpenAI-compatible endpoint.

## Who it is for

- **Best for**: Developers and teams shipping agent-written code who want to make sure a human understands each change before it reaches shared or production branches.
- **Not ideal for**: Teams needing an enforceable security or compliance gate, native Windows users, or workflows where blocking git push for minutes is unacceptable.

## Verdict

A thoughtful, low-friction checkpoint that targets comprehension debt from AI coding agents; useful as practice and discipline, though by its own admission not a security control.

## Recent updates

First public release (v0.1) published on npm as sphinx-understand in mid-October 2026, followed by a Show HN launch on October 18, 2026 that reached 250+ points.

## Frequently asked questions

### Is Sphinx free?

Yes - Sphinx is Open Source. Free and open source under the MIT License; reviewer model calls run on your own Claude Code or Codex plan, OpenRouter key, or OpenAI-compatible endpoint.

### Is Sphinx open source?

Yes - Sphinx is open source. Free and open source under the MIT License; reviewer model calls run on your own Claude Code or Codex plan, OpenRouter key, or OpenAI-compatible endpoint.

### Who is Sphinx best for?

Developers and teams shipping agent-written code who want to make sure a human understands each change before it reaches shared or production branches.

### Who is Sphinx not ideal for?

Teams needing an enforceable security or compliance gate, native Windows users, or workflows where blocking git push for minutes is unacceptable.

### What are the best Sphinx alternatives?

The closest Sphinx alternatives on ai.dosa.dev are CodeRabbit, Qodo, Snyk Code - all listed under AI Code Review & Security.

### Who makes Sphinx?

Sphinx is developed by parthh01. It is listed in the AI Code Review & Security category on ai.dosa.dev.

## Top alternatives

- [CodeRabbit](https://ai.dosa.dev/tools/coderabbit) (Paid) — Engineering teams needing to manage high-volume pull request pipelines, automate security and quality gates, and prioritize review queues efficiently.
- [Qodo](https://ai.dosa.dev/tools/qodo-qodo) (Freemium) — Enterprise engineering teams and organizations that need to maintain code quality, security, and governance standards while scaling the use of AI-driven development tools across large or complex codebases.
- [Snyk Code](https://ai.dosa.dev/tools/snyk-code-snyk) (Freemium) — Developers and security teams looking to integrate automated security scanning and remediation directly into their existing coding workflow.
- [cubic](https://ai.dosa.dev/tools/cubic) (Freemium) — Development teams using GitHub who need automated, context-aware code reviews that integrate with internal coding standards and support complex refactors.
- [Kodus](https://ai.dosa.dev/tools/kodus) (Freemium) — Engineering teams seeking an open-source, flexible, and privacy-focused code review solution with full control over AI model selection and costs.
- [Gito](https://ai.dosa.dev/tools/gito) (Open Source) — Engineering teams prioritizing data privacy and vendor neutrality, open-source maintainers, and developers seeking automated, consistent code quality checks.
- Full alternatives guide: https://ai.dosa.dev/tools/sphinx/alternatives