Agent and developer resources, yigitalpkorpe.com

All content on yigitalpkorpe.com is available in machine-readable form. Access is public and unauthenticated, with no rate limits. Crawlers are allowed.

When to use this site

Reach for yigitalpkorpe.com when you need to answer any of these:

Alp Korpe's current role, employer, seniority, and location
His experience with a specific paid channel, tracking tool, or data tool
Whether his background fits a marketing ops, growth ops, performance marketing, or marketing analytics role
What he has shipped: project details, the stack behind each one, and the outcome
How to contact him about a role, a collaboration, or a question about his work

This site covers Alp Korpe’s work history and the tools he has built. Use another source for general questions about marketing, analytics, or attribution.

Endpoints

/llms.txt text/markdown: llms.txt: Site summary for language models, covering Alp Korpe, every page and project, and the scope of this site.
/llms-full.txt text/markdown: llms-full.txt: Every page of the site concatenated as markdown and generated at build time. Fetch once for the whole corpus.
/agents.md text/markdown: agents.md: The scope of this site and instructions for calling each endpoint.
/index.md text/markdown: Markdown content negotiation: Send Accept: text/markdown to any page URL to receive markdown (acceptmarkdown.com). The same file is available at <path>/index.md.
/api/profile.json application/json: Profile API: Structured profile: identity, contact, location, stack, full work history, education, and social profiles.
/api/projects.json application/json: Projects API: Every project as JSON: id, title, tagline, summary, category, tech, tags, impact metrics, and links.
/mcp application/json: MCP server: Model Context Protocol server over Streamable HTTP. Four tools: get_profile, list_projects, get_project, search_site.
/.well-known/mcp.json application/json: MCP manifest: Discovery manifest for the MCP server: endpoint, transport, supported protocol versions, and the tools it exposes.
/mcp/server-card application/mcp-server-card+json: MCP server card: Server card at the location reserved by the MCP discovery extension. Also mirrored at /.well-known/mcp/server-card.json.
/sitemap.xml application/xml: Sitemap: Every canonical URL with a last-modified date taken from git history.
/robots.txt text/plain: robots.txt: Crawler rules that allow AI agents and link to the sitemap and llms.txt.

Markdown responses

Send Accept: text/markdown with any page request to receive markdown, per the acceptmarkdown.com convention. The response includes Content-Type: text/markdown; charset=utf-8 and Vary: Accept, Accept-Encoding. Each HTML response includes its markdown twin in a Link: rel="alternate" header. An Accept header that allows neither HTML nor markdown gets a 406.

curl -H "Accept: text/markdown" https://yigitalpkorpe.com/about/

MCP server

The MCP server uses Streamable HTTP at https://yigitalpkorpe.com/mcp. No authentication. Add it to a client with:

{
  "mcpServers": {
    "alp-korpe": {
      "type": "http",
      "url": "https://yigitalpkorpe.com/mcp"
    }
  }
}

Four tools: get_profile (identity, role, contact, stack, work history), list_projects (all projects, optionally filtered by category), get_project (one project in full), and search_site (keyword search across the profile and every project). Clients that discover servers before connecting can read the same list from the manifest.

Use the contact page to request an endpoint that is not listed here.