Podcast Guest Crm

What the podcast-guest-crm MCP server does

The podcast-guest-crm MCP server is intended for managing the operational work around booking podcast guests. Its listing describes five MCP tools covering three areas: moving guests through a booking pipeline, preparing outreach drafts, and viewing analytics. The repository models a six-step lifecycle: Discover, Outreach, Scheduled, Recorded, Published, and Follow-up.

The surrounding application includes guest records with contact details, company and role information, biography, topics, social links, priority, and lifecycle stage. It also provides a dashboard, a kanban-style pipeline, guest detail views, notifications, and analytics. These application features give context for the MCP workflow, although the supplied material does not provide the individual MCP tool names or their parameter schemas.

How it works

The project is a TypeScript monorepo using Next.js for the web application and Fastify for the API. Drizzle ORM is used with SQLite or Turso, and the repository also describes Supabase for production deployment. The local development command starts a web interface on port 3000 and an API on port 3001; API documentation is available through Swagger UI at /docs.

AI functionality is centralized in the packages/ai package and uses Anthropic's Claude model. Documented AI operations include personalized outreach emails, guest-fit scoring, interview briefs, topic tagging, follow-up sequences, and social posts. Outreach generation can stream a live preview and also return structured output containing a subject, message body, confidence score, and reasoning. The MCP listing specifically calls out outreach drafts and analytics, so those broader product capabilities should not be assumed to be exposed as separate MCP tools.

Setup and configuration

For local development, clone the repository, install dependencies, and start the development process:

git clone https://github.com/RudrenduPaul/podcast-guest-crm
cd podcast-guest-crm
pnpm install
pnpm dev

The first local boot uses seed data containing 34 guests across all six stages. No environment variables are needed for this seeded development mode. The README states that this zero-configuration behavior does not apply to production. A production deployment requires Supabase and Anthropic credentials, and the environment validation schema stops the server during startup when a required secret is absent. The supplied material does not list the exact variable names or an MCP client configuration example.

Tools and capabilities

The five MCP tools are described at a category level as supporting:

• Managing guest movement through the podcast booking pipeline.
• Creating or working with outreach drafts.
• Accessing podcast booking analytics.

The repository also documents lifecycle validation, outreach activity tracking, conversion metrics, stage and topic charts, and a 12-week outreach timeline in the application. Confirm the repository's current MCP schemas before building an integration that depends on exact arguments, return fields, or write behavior.

Limitations and notes

The provided README excerpt does not name the five MCP tools, document their transport, show client configuration, or specify their request and response schemas. It also does not establish compatibility with a particular MCP desktop or code client. Local seeded data is suitable for development, but production operation depends on external Supabase and Anthropic credentials. Analytics are described as able to run without a backend dependency in the application, which may not describe the behavior of the MCP interface itself.