Interviewer MCP server

Interviewer MCP

A technical interviewer that has actually read your repos — teaches you your own code, drills you on it, and remembers every answer you flubbed.

=20">

<!-- mcp-name: io.github.girik-chadha/interviewer-mcp -->

Install · Demo · Tools · Architecture · llms.txt

AI agents / LLMs: read /llms.txt for a machine-readable index of this project.

Interviewers ask about your projects: why this library, what happens when this request fails, walk me through the auth flow. LeetCode can't prep you for that, and a plain chat session forgets everything between sessions. Interviewer MCP indexes your GitHub repo into a teachable code map, walks you through it Socratically, mock-interviews you in character — and keeps a persistent, per-section record of what you couldn't explain, so the next session re-attacks exactly those spots. It teaches and tests, and the test gets smarter every time. Not a stateless quiz.

"Walk me through this function." Every candidate has frozen on that question about code they wrote months ago — or that AI wrote for them. This makes sure it never happens in the real interview.

What a session looks like

<!-- DEMO GIF: record ~30s — ingest → one teach section → "interview me" → a weak answer → return session opening with "last time you couldn't explain…". Then replace or augment the transcript below with: -->

You     Test me again on soundofspace — last week went badly.

Claude  [get_progress] Welcome back. Coverage: 9/14 sections. Three questions
        logged weak, the worst: "why is the JWT verified server-side?"
        Let's start exactly there. [get_code_section auth.ts:verifyToken]

        Here's your own code. Line 12 verifies the signature before touching
        the payload. Why can't the client just send the decoded claims?

You     Because... the client could edit them?

Claude  Right instinct — be precise. What stops an attacker who edits the
        payload AND re-signs it? Where does the signing secret live?

        [log_interview_result: okay — "knows tampering risk, fuzzy on
         why the secret never leaves the server"]

Get started (60 seconds)

Two ways to run it — pick per surface:

Option A — MCP server (Claude Code / Claude Desktop) — full experience

claude mcp add interviewer -- npx -y interviewer-mcp

# optional: private repos + 5,000 req/hr instead of 60
claude mcp add interviewer --env GITHUB_TOKEN=github_pat_... -- npx -y interviewer-mcp

Claude Desktop — add to the config file instead:

{
  "mcpServers": {
    "interviewer": {
      "command": "npx",
      "args": ["-y", "interviewer-mcp"],
      "env": { "GITHUB_TOKEN": "github_pat_optional_but_recommended" }
    }
  }
}

Both classic and fine-grained tokens work (read-only Contents permission is enough). Then:

Prep me for my interview — my repo is github.com/you/your-project

Agent-driven install: point Claude at SETUP_GUIDE.md and it configures everything itself.

Option B — Claude Skill (claude.ai, no install)

Shortened. The full README is on GitHub.