mcp-maigret
An MCP server that runs the maigret OSINT tool in Docker: the agent checks where a username is taken and gets a report in the format you choose
- 267 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- The server launches Docker containers on your machine at the agent's command
- Requests to many third-party sites come from your IP, and some sites may rate-limit or block it
- Reports collect information about people's accounts, so a lawful basis and the account owner's consent matter
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/mcp-maigretDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The server launches Docker containers on your machine at the agent's command
- Requests to many third-party sites come from your IP, and some sites may rate-limit or block it
- Reports collect information about people's accounts, so a lawful basis and the account owner's consent matter
This is third-party code. Review the repository files before installing.
What it does
mcp-maigret is a TypeScript MCP server that runs maigret on an agent's request, an OSINT tool for finding a username across hundreds and thousands of sites. Maigret itself runs in a Docker container. The `soxoj/maigret` image is pulled on the first search, about 800 MB, so besides Docker and Node.js nothing needs installing. There are two tools. `search_username` checks a handle on the 500 most popular sites in the database, or on all of them with `use_all_sites`, and accepts a tag filter such as coding or photo. `parse_url` parses a profile URL, extracts data from the page and looks for linked usernames. The report is saved to the directory from `MAIGRET_REPORTS_DIR`. The txt, csv, json, html and xmind formats work. Input from the agent is treated as untrusted: usernames, tags and URLs go through strict validation, and Docker is launched through `execFile` with an argument array and no shell. The author asks you to use the tool only for lawful, authorized research on public data and to follow the law and each site's terms.
Who it is for. For OSINT and security professionals who check from Claude Code or Claude Desktop where a username is taken: their own, an organization's in an authorized assessment, or a public one.
Good fit when
- You want to see which sites hold your handle to find traces of your own accounts
- An authorized assessment needs the public accounts of an organization under one handle
- You have a profile link and need to find linked usernames
- You need an HTML, CSV or JSON report that the agent saves to disk for further analysis
- You want to narrow the search by site topic using tags, for example coding
Not a fit when
- Docker is not installed on the machine: maigret will not start without it
- You want to check a private person without a lawful basis: the author allows only lawful, authorized use
- You need a PDF report: with the `soxoj/maigret` image the file is not created
- The client or environment is untrusted: the server runs Docker and sends requests to sites from your IP
Example request
Use maigret to check where the handle octocat is taken, only on sites tagged coding, and save the report as HTMLMCP
- Transport
- stdio
- Authentication
- not required
Environment variables
MAIGRET_REPORTS_DIRrequired- Directory where the server saves maigret reports
Limitations
You need Node.js 18 or newer and a running Docker. The author developed and tested the server on macOS; Linux and Windows with Docker Desktop should work but are untested. The `MAIGRET_REPORTS_DIR` variable is required, and the directory is created if missing. The first search pulls the image, so it is slow. Some sites rate-limit or block automated requests. Format statuses were checked against maigret 0.6.6 and the latest image in October 2026; the maigret command line changes between releases, so results may drift. On Linux the user must be in the docker group.
How to disable. Remove the server from the client config, in Claude Code that is `claude mcp remove maigret`. `docker rmi soxoj/maigret` removes the image. Saved reports sit in the directory from `MAIGRET_REPORTS_DIR`; delete them by hand.
FAQ
How do I install mcp-maigret in Claude Code?
Set the reports directory and run `claude mcp add maigret -e MAIGRET_REPORTS_DIR="$HOME/maigret-reports" -- npx -y mcp-maigret`. For Claude Desktop the same block goes into `claude_desktop_config.json`, then restart the app.
Which report formats work?
txt by default, csv, json, html and xmind. For pdf maigret needs an optional extra that the `soxoj/maigret` image lacks, so the file does not appear and the tool says so.
Why is the first search so slow?
The server pulls the maigret image of about 800 MB. Later searches are faster, especially if you narrow them with tags and leave `use_all_sites` off.
How does `parse_url` differ from `search_username`?
`search_username` looks for a given handle on sites. `parse_url` takes a profile URL, extracts information from it and looks for linked usernames, and returns the result as maigret's output text.
Is it safe to give an agent this server?
Input is validated strictly and Docker runs without a shell. Still, the server runs containers on your machine and sends requests to many third-party sites from your IP, so connect it only to agents and clients you trust.
Discussion
No questions or reviews yet.
Related
MCP serversAn OSINT AI agent with an interactive mode, an MCP server and a CLI: about twenty tools to check domains, addresses, hashes and breaches
Devil's Eye: OSINT with a Russian layer
glaz-dyavola
An OSINT tool for passive reconnaissance: EGRUL, FSSP, VK, Telegram and Yandex alongside secret scanning and a 15-tool MCP server
A bridge between an MCP client and a Kali Linux terminal so the agent can run nmap, sqlmap, hydra and other tools for authorized pentesting and CTF
MCP server for the ARL-Next platform: returns ports, fingerprints, certificates, vulnerabilities and leaks for a domain, IP or URL from your asset database
MCP-Shield finds your MCP configs, connects to servers and scans tool descriptions for hidden commands, exfiltration channels and tool shadowing
An MCP server that gives the agent vulnerability triage: NVD data, EPSS, the CISA KEV catalog, MITRE ATT&CK mapping and one-call CVE risk scoring
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| mcp-maigret (this entry) | no | not required | 267 | |
| OpenOSINT | no | 1,831 | a standalone OSINT agent with about twenty tools for checking domains, addresses and breaches | |
| claude-osint | no | 2,881 | a skill set with OSINT methodology that does not run maigret |
README in short
The README opens with a lawful-use warning, then covers features, the architecture, a quick start for Claude Code, Claude Desktop and Smithery, tool parameters, a report-format table with statuses, security measures and a table of common errors. Detailed docs live on a separate site.