Log in
← Search assignments

MCP Server Documentation

Model Context Protocol API for AI Assistants

What is MCP?

The Model Context Protocol (MCP) is an open standard that allows AI assistants to connect to external data sources. Our MCP server lets AI tools like Claude, Gemini CLI, and Cursor search consulting assignments worldwide directly.

Instead of manually searching assignment listings, you can ask your AI assistant to find relevant consulting work for you.

Getting Started

Connect your AI assistant using our server URL:

https://mcp.consultant.dev/mcp

Gemini CLI extension

Install the public Consultant Assignments extension:

gemini extensions install https://github.com/Bytelope/consultant-jobs-plugin

Restart Gemini CLI after installation. Gemini discovers the hosted MCP server automatically and opens Consultant sign-in when authentication is required.

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "consultant-jobs": {
      "url": "https://mcp.consultant.dev/mcp"
    }
  }
}

Claude.ai

In Claude.ai settings, add the MCP server URL. You will be prompted to sign in with your Consultant account when connecting.

Cursor

Add to your Cursor MCP settings:

{
  "mcpServers": {
    "consultant-jobs": {
      "url": "https://mcp.consultant.dev/mcp"
    }
  }
}

VS Code (Copilot)

Add to settings.json:

{
  "mcp.servers": {
    "consultant-jobs": {
      "url": "https://mcp.consultant.dev/mcp"
    }
  }
}

Available Tools

search_assignments

Search for consultant assignments with filters.

  • query - Search query for assignment title, description, or skills
  • location - City or region (e.g., Stockholm, Berlin, London, Remote)
  • role - Filter by job role
  • seniority - junior, regular, senior, or lead
  • skills - Array of required skills
  • employment_type - Filter by employment type, e.g. contractor
  • includeRemote - Include assignments marked as remote (default true)
  • sort - quality, relevancy, posted-desc, posted-asc, deadline-asc, or deadline-desc
  • page - Page number, starting at 1
  • limit - Results per page (max 20)

get_assignment

Get full details of a specific assignment by its ID.

  • id - The assignment ID

upload_cv

Save a PDF CV to your account and start matching it against open assignments. Replaces any CV already saved on your account.

  • document_base64 - The CV file (PDF) encoded as base64
  • filename - The file name, e.g. cv.pdf
  • cv_text - Optional plain text of the CV, if the assistant has already read it

my_matches

List the assignments matched to your saved CV, with scores and links.

  • limit - Number of matches to return (max 25, default 10)

refresh_matches

Re-score your saved CV against assignments published since the last run. Manual refreshes may be limited to once every 24 hours; the tool reports when the next refresh is available.

Authentication

All API tools require OAuth authentication. When you first connect your AI assistant to the MCP server, you will be prompted to sign in with your Consultant account.

The CV tools also require that you have accepted AI processing of your CV on Consultant.dev. You can see and disconnect connected assistants under Connected apps on your account page.

Example Usage

Once connected, you can ask your AI assistant things like:

  • "Find backend development assignments in Stockholm"
  • "Show me senior React contractor assignments in Berlin, newest first"
  • "Get details for assignment ID abc123"
  • "Upload my CV and show me my five best matches"
  • "Refresh my matches"

Contact

Questions or issues? Contact us at contact@consultant.dev