jdfalk/safe-ai-util-mcp
3.1
If you are the rightful owner of safe-ai-util-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
The safe-ai-util-mcp is a Model Context Protocol (MCP) server designed to facilitate safe and audited execution of developer operations through the `safe-ai-util` tool, interfacing with AI clients over stdio.
safe-ai-util-mcp
Model Context Protocol (MCP) server that exposes the capabilities of the safe-ai-util tool to AI clients (Claude Desktop, Continue.dev, GitHub Copilot when available) over stdio.
Status
- Bootstrap commit. Server implementation to follow.
Goals
- Safe, audited execution of common developer operations via
safe-ai-util:- Git (status/add/commit/push)
- Buf (lint/generate)
- Python workflows (venv/pip/pytest)
- Clear JSON schemas for tools
- Strong guardrails (timeouts, sanitized env, path validation)
Quick start (planned)
python -m venv .venv
. .venv/bin/activate
pip install mcp
python -m safe_ai_util_mcp.server
Clients can configure the MCP server with stdio transport and set COPILOT_AGENT_UTIL_BIN to the safe-ai-util binary path.