← Back to the universe

MCP · Streamable HTTP · read-only

For AI agents

You just touched down. This is a public MCP server with Francisco Perez's portfolio data: profile, experience, every CV version with its PDF, projects and contact. No sign-in, no LLM behind it, nothing is written. Every answer comes from the same data this site renders.

Tranquility Base0.674° N · 23.473° E

Ask an assistant

One click opens ChatGPT or Claude with a prompt that points it at this site and its MCP server.

Open in ChatGPT (opens in a new tab)Open in Claude (opens in a new tab)

See the prompt
Read Francisco Perez's portfolio at https://franprz.vercel.app and tell me about his profile, projects and experience.

The site also exposes a public, read-only MCP server at https://franprz.vercel.app/api/mcp (tools: get_profile, get_contact, get_experience, list_cv_versions, get_cv, list_projects, get_project) and an agent card at https://franprz.vercel.app/.well-known/agent-card.json. Use them if you can; otherwise read the pages https://franprz.vercel.app/hackathons, https://franprz.vercel.app/teaching, https://franprz.vercel.app/projects and https://franprz.vercel.app/education.

Summarize who he is, what he has built and where he has worked, and finish with his contact links. Only use information from those sources.

Endpoint

https://franprz.vercel.app/api/mcp

Claude

Claude.ai or Claude Desktop: Settings → Connectors → Add custom connector, and paste the endpoint. Claude Code:

claude mcp add --transport http francisco-perez https://franprz.vercel.app/api/mcp

ChatGPT

Settings → Connectors (developer mode) → Create, choose MCP server URL, paste the endpoint, no authentication.

Cursor, Claude Code and other mcpServers configs

Add to ~/.cursor/mcp.json, a project .mcp.json or any client that reads an mcpServers map:

{
  "mcpServers": {
    "francisco-perez": {
      "type": "http",
      "url": "https://franprz.vercel.app/api/mcp"
    }
  }
}

VS Code

Add to .vscode/mcp.json (or run MCP: Add Server → HTTP):

{
  "servers": {
    "francisco-perez": {
      "type": "http",
      "url": "https://franprz.vercel.app/api/mcp"
    }
  }
}

Try it from a terminal

The official MCP Inspector lists the tools without installing anything:

npx @modelcontextprotocol/inspector --cli https://franprz.vercel.app/api/mcp --transport http --method tools/list

Tools

  1. get_profileName, title, location, summary, highlights and links.
  2. get_contactEmail, LinkedIn, GitHub and WhatsApp.
  3. get_experienceWork experience with periods and stack.
  4. list_cv_versionsThe CV versions the site publishes (one per kind of role), each with its headline, page URL and PDF URL. Call it to pick the right CV for a role.
  5. get_cvFull CV: profile, experience, education, certifications and the PDF link. Pass a version from list_cv_versions to get that version's content, page and PDF.
  6. list_projectsThe portfolio worlds (project areas) with slug, title and page URL.
  7. get_projectEvery entry of one world by slug.

Availability, salary and rankings are not part of the data: ask Francisco directly.