0x4D Model Context Protocol

MCP server for AI assistants

Ask your AI assistant directly about my profile, projects, availability and articles. The server is public, free and read-only.

Endpoint: https://0x68.de/mcp
Transport: Streamable HTTP (stateless, JSON)
Authentication: None
Access: Read-only, public content of 0x68.de

What is this?

The Model Context Protocol (MCP) is an open standard that lets AI assistants such as Claude use external tools and data sources. Through this server, your assistant can query the content of this website directly and cite it with a source link, instead of relying on outdated training data.

The server is also an example of my work: a small, secured MCP interface on Google App Engine, with restricted permissions, validated input and rate limiting. That’s exactly what I build for companies that want to open their own data to AI assistants safely.

Limits

Read-only. The server cannot send messages, book appointments or store anything. For enquiries, use the contact form or email.
Public content only. Answers come from the same text as this website.
Rate limit: 60 requests per minute per IP address, input up to 200 characters per search field.

Privacy

When you call it, Google Cloud (hosting) and Cloudflare process technically necessary data such as IP address, time and requested address, as with every page view. The server itself stores no requests; the IP address is only held briefly in memory for rate limiting. Read more in the privacy policy.

Setup

Claude Code

$ claude mcp add --transport http tim-fiedler https://0x68.de/mcp

Claude (web and desktop)

Settings → Connectors → Add custom connector, then enter this URL:

https://0x68.de/mcp

Cursor, Windsurf and other clients using an mcpServers config

{
  "mcpServers": {
    "tim-fiedler": {
      "url": "https://0x68.de/mcp"
    }
  }
}

VS Code (.vscode/mcp.json)

{
  "servers": {
    "tim-fiedler": {
      "type": "http",
      "url": "https://0x68.de/mcp"
    }
  }
}

Test with the MCP Inspector

$ npx @modelcontextprotocol/inspector

Plain HTTP

$ curl -s https://0x68.de/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_availability","arguments":{"language":"en"}}}'

Available tools

`get_profile`

Who I am, role, location, focus, education, languages and skills.

Parameters: language ('de' | 'en')

`list_services`

My services with the situation each addresses and reference projects.

Parameters: language ('de' | 'en')

`search_projects`

Client projects with role, industry, period and technologies, filterable by keyword or technology.

Parameters: query (string), technology (string), language ('de' | 'en')

`search_blog`

Blog articles, newest first, filterable by keyword and topic.

Parameters: query (string), tag (string), language ('de' | 'en'), limit (integer)

`get_article`

The full text of one article as Markdown, with key takeaways and FAQ.

Parameters: slug (string), language ('de' | 'en')

`get_availability`

When and where I'm available, on site and remote.

Parameters: language ('de' | 'en')

`get_contact_options`

Email, contact form, response time and PGP key.

Parameters: language ('de' | 'en')

Example questions for your assistant

“Which projects has Tim Fiedler built with Rust?”
“Is Tim Fiedler available for a project in Bremen in February 2027?”
“What experience does he have with RAG on sensitive data? Summarise his article on it.”
“How do I reach him, and how do I encrypt my message to him?”