for ai assistants

Ask your AI about me

More people now ask an AI assistant to research someone before a call. Connect yours to this site and it reads my projects, career, posts and interview answers straight from the source, instead of guessing.

The server

This site runs an MCP server: a standard plug that AI apps like Claude, ChatGPT, Cursor and VS Code understand. It's free, read-only and needs no sign-in.

https://mcp.himanshuverma.dev/mcp

Claude (web and desktop apps)

Open Customize → Connectors.
Click + Add, then Add custom connector.
Name it Himanshu Verma, paste the server address above and click Continue.
Choose No sign in, then click Add.

Free plans can add one custom connector. On Team and Enterprise plans, an owner adds it for the organisation first. Menu names may differ slightly between app versions.

Claude Code

claude mcp add --transport http himanshu https://mcp.himanshuverma.dev/mcp

Cursor, VS Code and other MCP apps

Cursor (~/.cursor/mcp.json):

{
  "mcpServers": {
    "himanshu": {
      "url": "https://mcp.himanshuverma.dev/mcp"
    }
  }
}

VS Code (.vscode/mcp.json):

{
  "servers": {
    "himanshu": {
      "type": "http",
      "url": "https://mcp.himanshuverma.dev/mcp"
    }
  }
}

ChatGPT and other apps that accept custom MCP connectors (in ChatGPT, through developer mode): add a connector with the same address and no authentication.

Things to ask

“Using the Himanshu Verma connector, how well does he fit this job description? [paste it]”
“What has Himanshu built with MCP and LangGraph? Link the code.”
“Quiz me on RAG with questions from Himanshu's interview Q&A.”
“Summarise his blog post on vectorless RAG in five bullet points.”

What it can look up

Tool	What it returns
`get_profile`	Who Himanshu is: role, summary, skills, certifications, what he's open to, links.
`get_career_history`	His career, job by job, with what he built and the results.
`list_projects`	His projects, optionally filtered by a tag like agents or nlp.
`get_project`	One project in full: what it does, the stack, code snippet, repo and demo.
`list_blog_posts / read_blog_post`	His blog posts, and the full text of any one of them.
`list_interview_questions / get_interview_answer`	His interview Q&A by topic or difficulty, and the full answers.
`get_recommendations`	What colleagues say: his LinkedIn recommendations, word for word, shared with their permission.
`search`	Searches everything above at once.
`get_contact_info`	How to reach him: email and LinkedIn (no phone number, on purpose).

Only what's already public on this site. It can't change anything, it doesn't see who you are, and it has no phone number to give out.

No AI app? Use your terminal

curl -sL himanshuverma.dev

And for language models and crawlers: llms.txt (a map of the site), llms-full.txt (everything as one Markdown file) and profile.json (the same, as data).