What Podcastindex MCP server does
Podcastindex MCP server connects an MCP-compatible agent to the Podcast Index API at podcastindex.org. It provides structured access to podcast records, feed metadata, recent episodes, and trending rankings. The server is suited to agents that need to identify shows, inspect a known feed, retrieve episode details, or discover podcasts by popularity.
The package exposes four primary tools. search_podcasts looks for shows by a text query and can limit the number of returned matches. get_podcast retrieves detailed information for a podcast using its Podcast Index feed ID. episodes lists recent episodes for that feed, including publication information, runtime, audio URL, and episode or season numbers. trending returns shows ranked by recent activity and accepts optional language and category filters.
How it works
The hosted option uses the Pipeworx gateway at https://gateway.pipeworx.io/podcastindex/mcp. A client can discover and call the Podcast Index tools through the normal MCP connection. The scoped gateway endpoint also exposes shared Pipeworx meta-tools, so its tool list is larger than the four Podcast Index tools. Those shared tools can route questions across the wider catalog; the connection's initialize response provides the authoritative scope for the current connection.
The full gateway is available at https://gateway.pipeworx.io/mcp when an agent needs all connected data sources listed directly. Both gateway URLs reach the same gateway, while the directly listed pack tools differ. The gateway also provides HTTP access, including POST requests to tool-specific endpoints such as podcastindex_search_podcasts.
For local use, the package runs as an MCP stdio process. In that mode, messages use stdin and stdout, and the server responds to MCP operations including initialization, tool listing, and tool calls. The local process exposes only the Podcast Index pack tools and does not add the gateway's shared meta-tools.
Setup and configuration
Podcastindex MCP server can be added to a client with the hosted URL:
{
"mcpServers": {
"podcastindex": {
"url": "https://gateway.pipeworx.io/podcastindex/mcp"
}
}
}
A local configuration can launch the package with npx:
{
"mcpServers": {
"podcastindex": {
"command": "npx",
"args": ["-y", "@pipeworx/mcp-podcastindex"]
}
}
}
The standalone mode does not require a Pipeworx account or a gateway round trip. The README also states that the first hosted calls do not require an account. No required environment variables or additional configuration values are documented.
Tools and capabilities
- Search the open podcast database by name or keyword.
- Retrieve show title, author, description, ownership, categories, language, artwork, and episode count.
- List recent episodes with titles, descriptions, dates, durations, audio URLs, and numbering.
- Find trending shows and optionally filter results by language or category.
- Call the hosted tools through MCP or HTTP, or run the package locally over stdio.
Limitations and notes
Podcastindex MCP server is focused on discovery and metadata retrieval. The provided material does not document tools for subscribing to shows, downloading audio, playing episodes, editing feeds, or publishing podcast content.
Results depend on the Podcast Index data exposed by the API. A podcast lookup requires the service's feed ID, which can be obtained from search results or another source. The gateway pack endpoint includes shared tools in addition to the pack-specific tools; use the local stdio mode if the agent should see only the four Podcast Index tools. The project is released under the MIT license.