About Izzet Yildirim
Izzet Yildirim's read-only MCP server lets AI agents retrieve information about his research, publications, CV, and availability from public website data. It requires no sign-in and can be used through a local bridge or hosted endpoint.
Key features
- 01Read-only access to public profile data
- 02Search publication catalog and paper text
- 03Retrieve CV and paper sections
- 04No sign-in required
- 05Local stdio-to-HTTP bridge
Use cases
- 01Find publications on a research topic
- 02Read sections of a published paper
- 03Review CV experience or skills
- 04Check availability and contact information
MCP support
Exposes profile, CV, publication, and paper-search tools to MCP clients. The package forwards requests to `https://iyildirim.com/api/mcp`, or clients can connect directly to the hosted endpoint.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://iyildirim.com/api/mcpclaude mcp add --transport http izzet https://iyildirim.com/api/mcp{"mcpServers":{"izzet":{"type":"http","url":"https://iyildirim.com/api/mcp"}}}Open in Cursor{"servers":{"izzet":{"type":"http","url":"https://iyildirim.com/api/mcp"}}}Open in VS Code{"mcpServers":{"izzet":{"type":"http","url":"https://iyildirim.com/api/mcp"}}}{"mcpServers":{"izzet":{"serverUrl":"https://iyildirim.com/api/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEiyildirim
An MCP server for Izzet Yildirim. Connect it to your AI agent and ask about his research, papers, CV, and availability.
It is read-only and needs no sign-in. It answers from the public data on iyildirim.com, including the full text of his published papers.
Connect
Claude Code
claude mcp add iyildirim -- npx -y iyildirim
Claude Desktop, Cursor, VS Code, Windsurf (any client that launches local servers): add to the MCP config.
{
"mcpServers": {
"iyildirim": { "command": "npx", "args": ["-y", "iyildirim"] }
}
}
No install: most clients can connect straight to the hosted server, which is what this package forwards to.
claude mcp add --transport http iyildirim https://iyildirim.com/api/mcp
Claude, ChatGPT, and Cursor also accept https://iyildirim.com/api/mcp as a custom connector URL.
Then ask
What has Izzet published on I/O tracing?
Read the system design section of the AgentStage paper and summarize it.
Is Izzet a fit for this role? (paste a job description)
Tools
| Tool | What it does |
|---|---|
get_cv_section | One section of his CV: experience, education, service, certifications, or skills |
get_paper_section | The text of one section of a public paper |
get_profile | Start here: overview, what he is looking for, contact, links |
get_publication | The catalog entry for one publication (summary, venue, abstract, links) |
list_paper_sections | The sections of a public paper |
search_paper_text | Search the full text of public papers for passages |
search_publications | Search the publication catalog by topic, system, or venue |
How it works
This package is a thin stdio-to-HTTP bridge. It holds no tools or data of its own: each message goes to https://iyildirim.com/api/mcp and the reply comes back unchanged, so the tools are always as current as the site. No dependencies.
Set IYILDIRIM_MCP_URL to point it at another endpoint, for example a preview deployment.
Develop
npm install
npm run validate # format check, types, tests, build
Release
Releases are published to npm, and then listed in the official MCP registry as io.github.izzet/mcp, by publish.yml when a version tag is pushed.
- Bump
versioninpackage.jsonand inserver.json(bothversionandpackages[0].version) in a pull request and merge it. A test and the workflow check that they match. - On
main:git tag v0.1.1 && git push origin v0.1.1(the tag must match the version).
Publishing needs either a trusted publisher configured for this repo and publish.yml on npmjs.com (no secret), or an NPM_TOKEN repository secret.
MIT licensed.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.