Public metadata for the klaro.legal MCP server: contracts and official letters explained in plain language

klaro.legal MCP server: contracts and official letters in plain language

Paste a contract, an official letter, a tax assessment or a set of terms andconditions, and get it back explained clause by clause in everyday language,from Claude, ChatGPT, Cursor or any MCP client. Text, a PDF or photos of thepages all work: a photographed letter is read by OCR before the analysis. Sixlanguages. No account, no API key. Hosted by klaro.legal;this repository holds the public metadata, the server code is not open source.

https://klaro.legal/api/mcp

Streamable HTTP · no authentication · listed in theofficial MCP registryas legal.klaro/document-explainer and as a.

What it explains, and what it does not

It explains what a document says. It does not review it, score it, flagclauses as unfair, or tell you what to do. That line is deliberate: klaro.legalis software, not a law firm. What matters in your case is for a qualified lawyerto decide, and every answer says so.

Typical documents: rental agreements, employment contracts, letters from a taxoffice or an immigration authority, insurance policies, terms and conditions,utility and telecoms contracts.

Tools

  • explain_document: send the document as text, as a base64-encoded pdf, oras base64-encoded images of its pages (JPEG, PNG, HEIC, WebP; up to 15 pagesand 8 MB decoded). Exactly one of the three. language picks the language ofthe explanation. Returns a short summary plus one entry per clause: theoriginal wording, then what it means. Plain text takes 20 to 60 seconds, ascan or photo longer because of OCR. read-only
  • get_explanation: if explain_document was still running, it returns ananalysis_id — pass it here to pick up the finished result. read-only

Explanation languages: de, en, es, fr, it, pt. The document itselfmay be in any language.

Setup

{
  "mcpServers": {
    "klaro-legal": {
      "type": "streamableHttp",
      "url": "https://klaro.legal/api/mcp"
    }
  }
}

See llms-install.md for a one-call check that it works.

Text, PDF or photos

Three ways in, one at a time:

  • text — the fastest route. Most clients already extract the text of a PDFby themselves, and then this is all that is needed.
  • pdf — the file itself, base64-encoded, up to 8 MB decoded. Used when thePDF has no usable text layer, or when its layout carries meaning.
  • images — photos or scans of the pages, base64-encoded, in page order, upto 15 pages and 8 MB decoded in total. This is the route for a letter someonephotographed with a phone: OCR reads it first, then the analysis runs.

Anything larger than 8 MB is better handled by theupload form, which takes files up to 50 MB — abase64 payload travels through the client's context window, and that is wherethe limit comes from, not from the analysis.

Privacy

Documents are anonymised before the analysis and processed inside the EU. Theanalysis itself runs on AWS Bedrock in Paris (eu-west-3); the text is not sentto any other AI provider. Details: klaro.legal/en-us/privacy

Also available

Operated by Martin Schenk S.L., Madrid · klaro.legal ·questions: [email protected]

MCP Server · Populars

MCP Server · New

    voygr-tech

    PlaceCall ☎️

    The phone is your last API. Give your agent a voice to call any US business - reservations, inquiries, quotes. Verified outcome + transcript. First 250 calls free.

    Community voygr-tech
    colibird-ai

    LMCP — Context and Actions for Your AI

    192+ local tools for Claude, ChatGPT, Cursor & Grok — Mail, iMessage, Teams, Slack, WhatsApp, OneDrive, Google Drive, Zoom, Outlook, Office. Native macOS and Windows, runs on your computer, no API keys.

    Community colibird-ai
    HoldMyBeer-gg

    blenderwright

    MCP server for Blender. 191 tools for modelling, materials, rendering and more, driven by Claude, Cursor, Codex, Ollama, llama.cpp, or any MCP client.

    Community HoldMyBeer-gg
    tesseron-dev

    tesseron

    Tesseron protocol, documentation, MCP gateway and Claude Code plugin. Created and maintained by Eigenwise.

    Community tesseron-dev
    getsentry

    sentry-mcp

    An MCP server for interacting with Sentry via LLMs.

    Community getsentry