GitGuardian
About
Scan projects for over 500 types of secrets using GitGuardian's API to prevent credential leaks.
Details
- Author
- gitguardian
- Categories
- Developer Tools, Security, Infrastructure, Other, API
- Tags
- #git
Jump to
Setup
Install GitGuardian in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/gitguardian/gg-mcp
Follow the installation instructions in the repository README, then restart your MCP client.
Bring GitGuardian's secret detection and incident management into your AI agent. Scan code for credentials before they leak, triage existing incidents, generate honeytokens, and remediate findings — all from inside your IDE or chat client, backed by GitGuardian's 500+ detectors.
[!CAUTION] MCP servers are an emerging technology. Agents act on your behalf and under your responsibility. Use trusted MCP servers and review agent actions when they interact with tools. To limit blast radius the server defaults to read-only-leaning permissions; what is actually exposed is determined by the OAuth scopes your access token holds.
- Secret scanning— proactively scan files for leaked credentials.
- Incident management— list, filter, assign, resolve, and tag incidents (both internal and Public Monitoring incidents).
- Honeytokens— generate honeytokens and list existing ones.
- Code-fix automation— open pull requests that remediate secrets in repositories your workspace monitors.
The exact set of tools exposed to your agent depends on the OAuth scopes granted to your access token.
Scan this codebase for any leaked secrets or credentials.
Remediate all incidents related to my project.
Check if there are any new security incidents assigned to me.
Help me understand this security incident and provide remediation steps.
Generate a new honeytoken for monitoring AWS credential access.
Create a honeytoken named 'dev-database' and hide it in config files.
The recommended way to run the GitGuardian MCP server is to point your MCP client at the hosted server. The MCP client handles OAuth automatically; no local install, no token to manage, nouvx.
Pick the URL that matches your GitGuardian region:
{ "mcpServers": { "GitGuardian": { "type": "http", "url": "https://mcp.gitguardian.com/mcp" } } }
Edit~/Library/Application Support/Claude Desktop/mcp.json(macOS) or%APPDATA%\Claude Desktop\mcp.json(Windows). Same JSON as Cursor. Claude Desktop versions that pre-date HTTP MCP support need theLocal stdio fallback.
Add the server inSettings → Connectors → Add custom connectorwith the URL above. OAuth is handled in the browser tab.
Edit~/Library/Application Support/Windsurf/mcp.json(or~/.config/Windsurf/mcp.jsonon Linux):
{ "mcp": { "servers": { "GitGuardian": { "type": "http", "url": "https://mcp.gitguardian.com/mcp" } } } }
Edit~/Library/Application Support/Zed/mcp.json(or~/.config/Zed/mcp.jsonon Linux) with the sametype: httpsnippet.
Two deployment paths are supported. Pick based on where your GitGuardian instance lives and what tradeoffs you accept.
Most users do not need to touch this — theQuick startconfig implicitly uses theOAuth proxymode on the hosted server, and theLocal stdioconfig usesPAT env.
There are four authentication modes the server can run in; you pick one via env vars.
[!NOTE] Browser-driven OAuth in stdio mode isdeprecated. New stdio deployments should authenticate with a PAT; OAuth-driven flows should use the hosted or self-hosted HTTP server. The stdio OAuth code path will be removed in a future release; until then it remains the default in stdio for backward compatibility.
For CI/CD, airgapped environments, or older MCP clients, run the server locally over stdio with a PAT:
{ "mcpServers": { "GitGuardian": { "command": "uvx", "args": [ "--from", "git+https://github.com/GitGuardian/ggmcp.git", "gg-mcp-server" ], "env": { "ENABLE_LOCAL_OAUTH": "false", "GITGUARDIAN_PERSONAL_ACCESS_TOKEN": "your_pat_here", "GITGUARDIAN_URL": "https://dashboard.gitguardian.com" } } } }
Create a PAT in your GitGuardian dashboard underAPI → Personal Access Tokens. The set of tools the server exposes depends on the PAT's scopes.
For Claude Desktop on macOS, thecommandfield needs theabsolute pathtouvx(e.g./Users/you/.local/bin/uvx) — Claude Desktop does not resolve$PATHfor MCP servers.
The MCP server will be soon available out of the box as part of your GitGuardian self-hosted deployment (Helm chart). This section is only meant to describe how it works, but you don't have to set it up.
A Docker image is published atghcr.io/gitguardian/mcp-server. Run it behind a reverse proxy that terminates TLS, then point your MCP clients at it. The container exposes the StreamableHTTP transport on port 8000 by default.
docker run -p 8000:8000 \ -e GITGUARDIAN_URL=https://dashboard.gitguardian.mycorp.local \ -e IS_ON_PREM=true \ -e MCP_BASE_URL=https://mcp.mycorp.local \ -e MCP_OAUTH_PROXY_ENABLED=true \ -e ENABLE_LOCAL_OAUTH=false \ ghcr.io/gitguardian/mcp-server:latest \ gunicorn --workers=4 --worker-class=uvicorn.workers.UvicornWorker \ -b 0.0.0.0:8000 gg_mcp_server.http_app:app
IS_ON_PREM=truetells the server it talks to a self-hosted GIM instance (API served under/exposed/v1, self-hosted scope set). When unset, the server guesses from theGITGUARDIAN_URLhostname, which fails for self-hosted instances deployed under agitguardian.com/gitguardian.techdomain — set it explicitly for any self-hosted deployment.
MCP_OAUTH_PROXY_ENABLED=truemakes the server advertise itself as an OAuth Protected Resource (RFC 9728) and proxy/authorize,/token,/registerto your GitGuardian dashboard. MCP clients then run the OAuth flow against your domain.
Thedeveloper-mcp-serverandsecops-mcp-serverconsole scripts are deprecated and re-export the unifiedgg-mcp-server. Update your MCP client configuration to invokegg-mcp-serverdirectly; both old scripts will be removed in a future release.
Have a use case that isn't covered?Open an issuewith your idea.
SeeDEVELOPMENT.mdfor contributing, running tests, and adding new tools.
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
Official managed MCP server for the Cycode platform. Exposes projects, violations, compliance frameworks, audit logs, brokers, SBOM, and more via OAuth-authenticated remote access — no local install required.
Provides SD Elements API integration for security and compliance management.
Interact with the Zeropath vulnerability management API.
The new Bright MCP Server enables AI assistants to directly interact with Bright and manage security scans end-to-end. Using natural language prompts, AI can now analyze existing coverage, identify missing or hidden entrypoints, add them to the project, select the most relevant security tests based on the application’s technology, and start scans automatically.
Securely access secrets from Doppler's secret management platform using a Doppler API token.
Boost security in your dev lifecycle via SAST, SCA, Secrets & IaC scanning with Cycode.
Enable AI agents to secure code with Semgrep.
An engineering governance and safety control plane for AI coding agents to enforce strict SDLC discipline, quality gates, and security branch protections.
Control Apache Airflow via its API using JWT authentication.
AI-powered security scanning. Scans code, files, and git diffs for vulnerabilities in real-time using the Armis scanning API.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.





