Show HN: Sphinx – A hook that quizzes you on your AI's code before shipping
Points and comments are a snapshot, not live.
Sphinx is a Git hook that quizzes developers on AI-generated code before allowing a push.
Sphinx is a free, open-source Git hook for agent-assisted development. It pauses pushes on protected branches and asks a question about what the code change does, particularly when things go wrong. It works with Codex, Claude Code, Cursor, OpenCode, and Pi. The developer must answer correctly to proceed; wrong answers trigger learning. A `sphinx yolo` command bypasses the check, and all skips are logged. It uses minimal tokens and runs on any OpenAI-compatible endpoint.
What commenters are saying
Commenters are sharply divided. Skeptics argue the tool adds friction for effective users and is useless for those who don't care, with some predicting AI agents will soon interact with Sphinx. Others defend it as a necessary check for teams pumping AI code without understanding, comparing it to old StackOverflow copying. Several note the name conflicts with Sphinx documentation. One commenter built a similar tool (vibecheck.dev) at the PR level. A few suggest it could be useful for OSS maintainers or imposing on careless colleagues.