# CC Safety Net

A pre-execution guard that blocks destructive Git and filesystem commands and access to secrets such as SSH keys before a coding agent's tool call actually runs.

CC Safety Net blocks destructive commands and access to secrets such as SSH keys and .env files before a coding agent's tool call runs, by parsing what a command actually does rather than matching text, so wrapping a command or reordering its flags does not slip past it; a broken configuration file never silently blocks nothing. It supports coding CLIs including Amp Code, Antigravity CLI, Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot CLI, Grok Build, Hermes Agent, Kimi Code, OpenClaw, OpenCode and Pi, on Windows, macOS and Linux, though Windows support for some of them is best effort.

It blocks known-destructive commands such as a hard Git reset, a forced push, or a recursive delete on a dangerous target, and blocks access to SSH keys, .env files, cloud credential folders and the credential files coding CLIs themselves keep, covering both the shell and an agent's own read, edit, write and search tools. Rules can be customized through a local GUI, extended with official rulebooks for Terraform, AWS, gcloud and Azure or a user's own JSON, and shared across a team by committing its policy folder to Git, with a status check flagging any project file that tries to loosen a stricter member setting. It is written in TypeScript and licensed under MIT.

Website: https://ccsafetynet.com