Consensus-OSINT MCP
One lookup by username, email or phone across 9,700+ checks from 19 OSINT projects, with reports, an evidence graph and an MCP server
- 232 ★ on the repository
- Last commit
- GPL-3.0
Broad access
Why this level
- Looking up other people's usernames, addresses and numbers collects personal data, and privacy law may require a legal basis
- Each check sends a request to an external service, and a wrong match harms an innocent person
- An assistant with this MCP runs lookups on its own, so restrict its tools and give only permitted targets
Install
- Works in
Manual install
python -m pip install "consensus-osint[modules,pdf]"Requires Python 3.10 or newer. Launch it with `consensus` after installing.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Looking up other people's usernames, addresses and numbers collects personal data, and privacy law may require a legal basis
- Each check sends a request to an external service, and a wrong match harms an innocent person
- An assistant with this MCP runs lookups on its own, so restrict its tools and give only permitted targets
This is third-party code. Review the repository files before installing.
What it does
Consensus merges the modules and site rules of 19 open-source OSINT projects into one catalogue of checks: more than 9,700 by the author's inventory of 16 October 2026. The `consensus user`, `consensus email` and `consensus phone` commands look for traces of a username, an email address or a phone number, and the program removes duplicate lookup routes so results do not repeat. Beyond found or not found, it shows a name, a profile bio and linked accounts where the source returns them. In interactive mode you set the search size: quick up to 300 checks, a custom limit or full. It also offers bulk search over a list, username variations, site and category filters, recursive follow-up on discovered identifiers and rechecking of earlier findings. Investigations are saved with notes. A report exports to HTML, PDF, CSV, JSON and other formats, and the evidence graph opens in a browser or exports to XMind and Neo4j. The MCP server starts with `consensus mcp --transport stdio` and gives an assistant the same username, email and phone lookups, paginated results and detailed evidence.
Who it is for. Open-source intelligence researchers, security specialists and journalists who check the digital footprint of their own or lawfully investigated usernames, addresses and numbers.
Good fit when
- A security specialist assesses where the company's work addresses and usernames surface online
- An investigation starts from one username and has to follow linked accounts and build an evidence graph
- A list of a hundred usernames must be checked in bulk with a summary CSV
- An assistant should run the lookup itself and summarize matches and inconclusive answers
Not a fit when
- You want to follow a person without consent or a legal basis: a match remains a lead and does not prove identity
- The data sits in private accounts or leaks: the tool queries only what external services return
- The result would go in as evidence unchecked: the author asks to verify important matches at the source
Example request
Use Consensus to search my username YOUR_USERNAME across up to 20 sources, then summarise the matches and inconclusive resultsMCP
- Transport
- stdio
- Authentication
- not required
Limitations
- Version 0.4.10 was tested on Windows and Linux, with no full test on macOS.
- A search contacts external services: some checks return errors, blocks or inconclusive answers, and no match does not prove an account is absent.
- Some modules depend on keys and provider access; the Settings, Setup / health screen lists missing keys.
- MCP installs as a separate `[mcp]` extra. Profile text it finds reaches the assistant as data, so limit the agent's tool permissions.
How to disable. Remove the `consensus` entry from `mcpServers` in the client settings and run `python -m pip uninstall consensus-osint`. Saved investigations live in the local reports folder.
FAQ
How many checks does Consensus have and where do they come from?
More than 9,700 by the 16 October 2026 inventory. The author collected them from 19 projects, including Maigret, User Scanner, WhatsMyName, Sherlock and Holehe, and states a weekly review of sources as a target.
How do I connect Consensus to Claude Code or another assistant?
Install the package with the `mcp` extra and add a server to the client config with the command `consensus` and arguments `mcp --transport stdio`. If the client cannot find the program, give the full path.
Can I use Consensus without installing Python?
Yes, Windows has an installer bundle in Releases: after `Install.cmd` Python and dependencies install themselves. macOS and Linux have `sh install.sh`.
Is a found profile proof of identity?
No. The author calls a match a lead, and an unavailable profile does not mean a deleted account. Important matches get verified at the source.
How do I update Consensus?
Close the program and run `consensus update`. `consensus update --repair` and `consensus update --rollback` handle repair and rollback.
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
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
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
Security Detections MCP: a detection rule database
Security Detections MCP
An MCP server that merges Sigma, Splunk, Elastic and KQL rules into one database for finding detections and ATT&CK gaps
A scanner that finds leaked keys, dangerous access and connected models in MCP configs of Claude Desktop, Cursor and VS Code.
Security scanner for MCP servers that looks for hardcoded keys, tool poisoning and injection flaws in repository code.
Comparison with similar tools
Cards that share the main search query, and how they differ.
README in short
The README opens by comparing Consensus with its 19 source projects by check count, duplicate share and update frequency. Features, screenshots, installation routes, updating and MCP setup follow. It closes with a warning that the findings are leads and do not count as proof.