softwareAI Security
AI Gatekeeper
See and control what AI coding agents do on your Windows PC, the way an antivirus controls programs.
Contact for price
Request a quoteKey features
- Live view of the files each AI agent reads and changes
- Block and allow lists by agent, folder and operation
- Hide from AI: agents can use a file while its secrets stay away from the model
- Pause or stop any agent or sub-agent
- Network kill switch to cut an agent off
- A record of what data was sent to AI
- Detection of unknown AI agents
- Alerts, HTML reports and a protection PIN
Overview
AI coding agents can reach the files on your PC, including passwords, keys and private documents. AI Gatekeeper is a Windows security tool that shows what each agent does, lets you decide what it may touch, and stops it when it crosses the line.
What you can do
- Watch agents live: see which files each agent and its sub-agents open, change, rename or delete.
- Set the rules: allow or block folders and files for each agent, and change the rules at any time.
- Hide secrets from AI: let an agent build and test a project while values such as those in a
.envfile are replaced before they reach the model. - Respond to violations: get an alert, pause the agent or stop it.
- Cut network access: block an agent from the network with one switch.
- Review what was shared: see which file contents were sent to AI, and which were hidden.
Start safely
Begin in Audit mode, which only records activity. Tune your rules, then switch to Enforce when you are ready.
Works with
Windows PCs running Claude Code, with integrations for GitHub Copilot, Cursor, OpenAI Codex, Gemini CLI, Windsurf, Cline and Antigravity.
Interested?
