Launching today

AI Access Auditor
Like a firewall prompt, but for AI access
1 follower
Like a firewall prompt, but for AI access
1 follower
Free app for macOS, Windows and Linux. See every permission, folder, login and key your AI agents (Claude Code, Codex, Gemini CLI, Cursor) have, and take any of it back.






Launch Team

AWS Activate CreditsCredits that fuel your startup’s next milestone
Promoted
Hi Product Hunt! 👋
I use several AI coding agents every day, and after a few months of clicking "Always allow" I had no idea what they could still do on my computer. Each agent keeps its own permissions in its own files: allowlisted commands, MCP servers with tokens, trusted folders, saved logins, hooks. Some even leave API keys in their session logs.
So I built AI Access Auditor. It:
- finds every permission Claude Code, Codex CLI, Gemini CLI, OpenCode, Cursor and VS Code have, and explains each one in a line;
- gives you a safety score, so you know where to start;
- revokes anything in one click, with a backup, so every change can be undone;
- pops up when an agent gains new access;
- can ask you before Claude Code, Codex or Gemini CLI runs a command.
It runs only on your computer: no account, no telemetry. The only network request is a daily update check you can turn off. It's free and open source (MIT), for macOS (signed and notarized), Windows and Linux.
This is my first public release, so I'd love your feedback. Which agents or settings should it cover next?