Discogs MCP

Discogs MCP plugs the entire Discogs music database into Claude. You get search across releases, artists, labels, and marketplace listings without leaving your editor or chat window. I've used it to price-check vinyl before buying, trace sample sources on old records, and pull together label rosters for research. It's straightforward: you ask Claude to find a release or artist, and it returns structured data from Discogs' API. The main appeal is speed. If you're already in Claude Desktop working on music journalism, crate-digging notes, or building a catalogue app, you skip the browser tab dance. The marketplace search is particularly useful for checking current prices on pressings, though you'll still need to visit Discogs to actually buy anything. Artist lookups pull discographies with release years and formats, which beats manually scrolling through Discogs' web interface when you're cross-referencing a dozen artists. You'll need a Discogs API token, which is free but requires an account. The setup is typical MCP fare: install via npx, drop the token in your environment config, restart Claude. Once it's running, queries feel natural. Ask for "1980s releases on Factory Records" or "marketplace prices for Aphex Twin Selected Ambient Works" and you get JSON-formatted results Claude can parse and summarise. Limitations are mostly about scope. This is read-only, so you can't add to your collection or place orders. The marketplace data is a snapshot, not live, so prices drift if listings change between queries. If you're not already a Discogs user or don't work with music metadata regularly, this won't create a use case for you. It's a tool for people who already lean on Discogs and want to fold it into their LLM workflow. I'd reach for this when building music databases, fact-checking liner notes, or researching reissues. It's less useful if you're just browsing casually, because the Discogs website is frankly better for that. But for structured queries and bulk lookups, it's a clean shortcut.