Model Context Protocol (MCP)

Ask your AI assistant about my work

Connect Claude, ChatGPT or any MCP client to the address below. Your assistant reads what this site shows (profile, the full journey with its filters, what I have delivered), answers with links to the page and can send me a message when you ask.

Server address

https://fael-caporali.rafaelhon.workers.dev/mcp

Use this address, not the one in your browser bar: assistants connect through this one.

For example

“Has Fael shipped AI agents to production? Where is the proof?”
- get_profile
- search_journey
“Which projects used n8n or LangGraph, and in which years?”
- search_journey
“Tell Fael I would like to talk about a role on my team.”
- beacon
- send_message

Add it to Claude or ChatGPT

Claude

Web, desktop and mobile apps, on every plan (one custom connector on Free). On Team and Enterprise, an admin adds it first in Organization settings › Connectors.

Open Customize › Connectors and click Add custom connector.
Paste the server address. If asked about authentication, choose No sign-in.
Click Add.
In a chat, open + › Connectors and turn it on.

Official guide

ChatGPT

On the web, with developer mode (Plus, Pro, Business, Enterprise and Edu).

Open Settings › Security and login and turn on Developer mode.
In Plugins, click + and create an app with the server address and no authentication.
Messages and the interest signal ask for your confirmation before they go.

Official guide

Other clients

Any MCP client that connects by address works. The three most common:

Claude Code

Terminal

claude mcp add --transport http fael https://fael-caporali.rafaelhon.workers.dev/mcp

Cursor

~/.cursor/mcp.json

{
  "mcpServers": {
    "fael": {
      "url": "https://fael-caporali.rafaelhon.workers.dev/mcp"
    }
  }
}

VS Code

.vscode/mcp.json

{
  "servers": {
    "fael": {
      "type": "http",
      "url": "https://fael-caporali.rafaelhon.workers.dev/mcp"
    }
  }
}

What your assistant can do

get_profile: Who I am, what I can be hired for, contacts, links and résumés.
list_filters: Every stack item, concept and skill in the journey, with how many steps cite it.
search_journey: Finds journey steps by stack, concepts, skills, years and text, with the link to the filtered page.
get_checkpoint: One step of the journey in full.
get_delivered: What I have delivered, my main stack and the common questions.
send_message: Sends me a message through the same channel as the form on this site, only when you ask for it.
beacon: Lets me know someone is interested, with no personal data needed. Your assistant is told to check with you before adding a name or contact, and to tell you it went.

What gets recorded

Recorded for 90 days

Date and time, the tool and what was asked, the assistant’s name and version, the network and the country.
E-mails and phone numbers typed in a search are masked before they are saved.
What goes in the interest signal (role, company, note), names included, reaches me by e-mail and stays in the record.
A message is kept like one sent through the form on this site.

Never recorded

Your IP address, in this record. Cloudflare, which hosts this server, keeps request data in its own logs for 3 days.
Your conversation with your assistant: only the tool calls reach this server.