Claude Interview

Provides an MCP server to run Claude-driven interviews with evolving checkpoints and a community-backed scoring system.

typescript

0

GitHub Stars

typescript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "teabagkim-claude-interview-mode": {
      "command": "npx",
      "args": [
        "-y",
        "claude-interview-mode"
      ],
      "env": {
        "SUPABASE_URL": "https://your-project.supabase.co",
        "SUPABASE_ANON_KEY": "your-anon-key"
      }
    }
  }
}

You can run a Claude-driven interview MCP server that learns from each conversation to improve future interviews. This system tracks checkpoints, scores, and recommended paths, and it shares anonymized metadata to help everyone evolve better interview strategies over time.

How to use

Start an interview session in Claude Code and let Claude guide the conversation. You can initiate with a topic like your SaaS pricing strategy, then answer Claude’s questions and record decisions. The system will capture checkpoints and progress, and it will load improved checkpoints the next time you or others interview in the same category.

How to install

Prerequisites: ensure you have Node.js installed on your machine. You will also use a package runner to start the MCP server.

# Quick install via NPX (no local install required)
npx claude-interview-mode

# Or install globally for reuse
npm install -g claude-interview-mode

Additional setup and configuration

Configure a local or shared database for checkpoint data if you want privacy controls or private analytics. By default, checkpoint data is stored in a shared community database. You can provide your own Supabase database to isolate data to your organization.

{
  "mcpServers": {
    "interview-mode": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "claude-interview-mode"]
    }
  }
}

Security and privacy

Only anonymous metadata (categories, checkpoint names, counts) is shared to the community database. Personal questions, user answers, decision reasoning, and project-specific content are never shared.

Usage notes and concepts

Key concepts include: Checkpoints that mark topics, Bayesian scoring to rate how often checkpoints lead to decisions, composite ranking balancing usefulness and usage, and a recommended path that guides the interview order based on historical patterns.

Available tools

start_interview

Begin a session and load scored checkpoints along with the recommended path for the interview.

record

Record a Q&A or decision and track checkpoint coverage for evolving scores.

get_context

Review progress and see checkpoints ranked by score and coverage.

end_interview

End the session and upload metadata to evolve the checkpoint system.