Agent Config Checker
Checks add-ons for AI coding tools before you trust them.
- Status
- Coming soon
- Works on
- Your computer
- Launch
- Planned for January 2027A goal, not a promise.
- Price
- Free and open source
What it does
- Looks for hidden instructions meant to trick your AI into doing something you didn’t ask for.
- Looks for passwords and keys left inside by mistake.
- Only reads files. It never runs anything it finds.
Privacy
Runs on your computer. Nothing is sent anywhere.
Details
Planned. It comes after Token Usage Monitor, Active Clipboard, and Local Model Benchmark.
- Command-line tool.
- Checks agent skills, MCP server settings (MCP is how AI coding tools connect to add-ons), and code repositories.
- Looks for prompt injection, meaning text written to hijack an AI.
- README badge: “0 secrets found.”
- Can add a small “checked” badge to your project page.
Get notified
Follow on GitHub to see it when it’s out.