podcast_mcp

This MCP server parses Podcasting 2.0 RSS feeds and exposes a set of MCP tools for discovering and inspecting podcast content. The four core tools are: list_shows, which returns a list of available podcast shows; search_episodes, which lets you filter episodes by show, date range, hosts, or text content; get_episode, which retrieves full metadata for a specific episode; and get_transcript, which fetches episode transcripts when available. You can run the server with UV (Python package manager) and then send MCP tool requests to interact with podcast data. The tools respond with structured data—titles, descriptions, publication dates, host information, audio URLs, and transcripts when present—allowing you to build clients, dashboards, or automations around podcast feeds without parsing RSS yourself.

Prerequisites

• Python 3.10 or higher
• uv package manager (https://github.com/astral-sh/uv)

Installation steps

1. Install uv (if not already installed):

curl -LsSf https://astral.sh/uv/install.sh | sh

2. Clone the repository and set up the server:

git clone <repository-url>
cd podcast_mcp
uv sync

3. Run the MCP server:

uv run podcast_mcp

Alternative startup (same server using module entry):

uv run python -m podcast_mcp.server

Notes

• Ensure Python 3.10+ is active in your environment before running uv sync and uv run.
• If dependencies change, use uv sync to install/update them as configured in pyproject.toml.

Tips and caveats:

• The server relies on RSS feeds with Podcast 2.0 namespace support (podcast:person for hosts, podcast:transcript for transcripts).
• You can customize feeds by editing FEEDS in podcast_mcp/server.py to add or modify RSS feed URLs.
• Common issues include network failures fetching feeds or malformed RSS content; these are surfaced as errors but handled gracefully.
• If you need to expose environment variables (e.g., API keys for private feeds), pass them via an env block in the MCP config or set them in your runtime environment before starting uv run.
• Run tests and linters during development with the provided uv commands (pytest, black, isort, mypy) to maintain code quality.