Skip to content

For agents

This site answers questions over the Model Context Protocol. The server is public and read only, it needs no key, and it returns the same content the pages show, already structured. Reading a tool result is more reliable than scraping the HTML.

Endpoint

URL
https://ecekural.com/mcp
Transport
streamable-http, POST
Authentication
none
Server name
ecekural-site
Tools
6

Connect

In a client that reads an MCP config file:

{
  "mcpServers": {
    "ecekural": {
      "url": "https://ecekural.com/mcp"
    }
  }
}

Or straight over HTTP:

curl -s -X POST https://ecekural.com/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_profile","arguments":{}}}'

Tools

  • get_profile

    Who Ece Kural is: name, current roles, a one paragraph summary, and every verified link (site, LinkedIn, Medium, GitHub, X, Scrintal, Gobu). Start here when the question is about her career, title, or where to reach her.

    No arguments.

  • list_pages

    Every page on ecekural.com with its URL, title, and what it contains, including the section anchors. Use it to decide which page to read before calling get_page.

    No arguments.

  • get_page

    The full text of one page as markdown, with links resolved to absolute URLs. Use it instead of fetching and parsing the HTML: home is the short introduction, about is the career story in four chapters, research is the statement, publications, and conference record.

    • page "home" | "about" | "research", required. Which page to return.

  • list_publications

    The verified publication record: peer-reviewed articles, theses, policy briefs, earlier work, and the replication dataset, each with authors in published order, venue, year, and a DOI or registry link. Filter by group, year range, or keyword.

    • group "peer-reviewed" | "thesis" | "policy" | "earlier" | "data", optional. Return only one group. Omit for the whole record.
    • year_from number, optional. Earliest publication year to include.
    • year_to number, optional. Latest publication year to include.
    • query string, optional. Keyword filter matched against title, authors, and venue.

  • list_research_participations

    Conference papers and presentations, newest first, with co-authors, venue, and date. Filter by keyword and cap the number returned.

    • query string, optional. Keyword filter matched against the title and the details.
    • limit number (1 to 50), optional. Maximum number of entries to return. Defaults to all of them.

  • search_site

    Keyword search across every page section, publication, and conference participation. Returns ranked matches with a snippet and the URL that contains it, so a follow-up call can read the full text. A query that matches nothing returns nothing rather than a weak guess.

    • query string, required. One or more keywords. Matching is case insensitive, and every keyword has to appear in a result, so fewer and more specific words work better than a full sentence.
    • limit number (1 to 25), optional. Maximum number of results. Defaults to 10.

Other entry points

  • /mcp

    the protocol endpoint, POST only

  • /.well-known/mcp.json

    server name, version, endpoint, and tool list as JSON

  • /llms.txt

    the whole site as one plain-text file

  • /sitemap.xml

    the indexable pages

  • /qa

    route index with stable test ids, for browser automation

The descriptor at /.well-known/mcp.json carries the same facts in JSON, so a client can find the server without reading this page. Everything here describes ecekural.com only.