Code Guardian runs on your own machine — the client starts it, so there is no endpoint to ping. 215 installs a week from pypi. Last commit 16 Oct 2026.
Safe to run? Will it work? Is it complete? A deterministic A-F grade for a repo, without running it.
Today is the operative word: we check Code Guardian every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add code-guardian -- uvx code-guardian-mcp
{
"mcpServers": {
"code-guardian": {
"args": [
"code-guardian-mcp"
],
"command": "uvx"
}
}
}
[mcp_servers.code-guardian]
command = "uvx"
args = ["code-guardian-mcp"]
{
"mcpServers": {
"code-guardian": {
"args": [
"code-guardian-mcp"
],
"command": "uvx"
}
}
}
{
"mcpServers": {
"code-guardian": {
"args": [
"code-guardian-mcp"
],
"command": "uvx"
}
}
}
This one needs environment variables set before it will start:
CODE_GUARDIAN_LLM_URL (Optional: local model for the explain tool (Ollama /api/chat or an OpenAI-compatible /v1/chat/completions URL)), CODE_GUARDIAN_LLM_MODEL (Optional: model name for the explain tool).
The author declared them in the registry entry; get the values from the project itself.
Deterministic code map for AI coding agents: a tree-sitter call graph with exact file:line over MCP.
Deterministic release gate for AI-written code and coding agents.
Deterministic layered code graph MCP server with framework overlays and SCIP interop.
Deterministic SAST/SCA findings + fix guidance for your GitHub repos, handed to your AI agent.
Deterministic AGENTS.md and repo analysis for coding agents, paid per call in USDC over x402.
Deterministic repository structure mapping for onboarding agents.
Run code an AI wrote without handing it everything you can reach.
Deterministic repository change-impact analysis for engineers and coding agents.
Answers built from our own checks of this server.