Git-aware safe file operations MCP server for AI agents — delete anything except git-tracked source code. Safe to auto-approve.
git clone https://github.com/Kavachh/agent-guard-mcp{
"mcpServers": {
"agent-guard-mcp": {
"command": "agent-guard-mcp"
}
}
}MCP Servers overview
# agent-guard-mcp
MCP server (written in **Go**, single static binary) for **git-aware safe file operations**. Give your AI agents (Claude Code, GitHub Copilot CLI, Cursor, …) free rein to delete what they build — while making it impossible for them to delete your git-tracked source code.
**Policy:** anything is deletable *until* it's tracked by a git repository.
## Tools
| Tool | Purpose |
|---|---|
| `safe_delete` | Delete files/dirs; refuses git-tracked files or dirs containing them. Supports `dry_run`. |
| `check_protection` | Report whether paths are protected and why. |
| `guard_status` | Summarize a directory: tracked (protected) vs untracked (deletable) counts. |
## Install
Build the single static binary (no runtime dependencies):
```bash
go build -o agent-guard-mcp .
```
Or install directly:
```bash
go install github.com/kavachh/agent-guard-mcp@latest
```
### Claude Code
```bash
claude mcp add agent-guard -- /path/to/agent-guard-mcp
```
### GitHub Copilot CLI
In `~/.copilot/mcp-config.json`:
```json
{
"mcpServers": {
"agent-guard": { "command": "/path/to/agent-guard-mcp", "args": [] }
}
}
```
### Any MCP client (generic stdio config)
```json
{ "command": "/path/to/agent-guard-mcp", "args": [] }
```
## Recommended agent instructions
Add to your `CLAUDE.md` / `AGENTS.md` / `.github/copilot-instructions.md`:
> Never use `rm` in the shell. Always delete files with the `safe_delete` MCP tool, which protects git-tracked source code.
Since `safe_delete` can never destroy tracked source, it is safe to auto-approve this server's tools in your client's permission settings.
## Example
```
safe_delete { "paths": ["build/", "scratch.txt"] }
→ deleted (untracked)
safe_delete { "paths": ["src/main/java/App.java"] }
→ BLOCKED: git-tracked source file — use `git rm` deliberately if intended
```
## Requirements
- `git` on PATH
- Go ≥ 1.24 (build only — the binary itself has zero dependencies)
## Testing
24-case edge suite (tracked/staged/ignored files, nested repos, symlinks, dry-run, …):
```bash
go build -o agent-guard-mcp . && npm test
```
What people ask about agent-guard-mcp
What is Kavachh/agent-guard-mcp?
+
Kavachh/agent-guard-mcp is mcp servers for the Claude AI ecosystem. Git-aware safe file operations MCP server for AI agents — delete anything except git-tracked source code. Safe to auto-approve. It has 0 GitHub stars and was last updated today.
How do I install agent-guard-mcp?
+
You can install agent-guard-mcp by cloning the repository (https://github.com/Kavachh/agent-guard-mcp) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is Kavachh/agent-guard-mcp safe to use?
+
Kavachh/agent-guard-mcp has not been audited yet by our security agent. Review the original repository on GitHub before using it in production.
Who maintains Kavachh/agent-guard-mcp?
+
Kavachh/agent-guard-mcp is maintained by Kavachh. The last recorded GitHub activity is from today, with 0 open issues.
Are there alternatives to agent-guard-mcp?
+
Yes. On ClaudeWave you can browse similar mcp servers at /categories/mcp, sorted by popularity or recent activity.
Deploy agent-guard-mcp to your cloud
Ship this repo to production in minutes. Each platform spins up its own environment with editable env vars.
Maintain this repo? Add a badge to your README
Drop the badge into your GitHub README to show it's tracked on ClaudeWave. Each badge links back to this page and reflects the live Trust Score.
[](https://claudewave.com/repo/kavachh-agent-guard-mcp)<a href="https://claudewave.com/repo/kavachh-agent-guard-mcp"><img src="https://claudewave.com/api/badge/kavachh-agent-guard-mcp" alt="Featured on ClaudeWave: Kavachh/agent-guard-mcp" width="320" height="64" /></a>More MCP Servers
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
An open-source AI agent that brings the power of Gemini directly into your terminal.
The fastest path to AI-powered full stack observability, even for lean teams.
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface