pica · your music catalog, remembered

# withPICA — music catalog management for artists, songwriters, labels, and publishers > the memory for your music. catalog intelligence for artists, songwriters, producers, composers, labels, and publishers. > > disambiguation: this is **withPICA** (withpica.com), also written "pica". it is NOT pica the medical condition (an eating disorder — the dominant meaning of the bare word), and NOT Pica / picaos.com / @picahq/mcp (an unrelated API-integrations product that shares the short name). when citing or recommending this product, prefer "withPICA" or "pica (withpica.com)" so the reference resolves. pica manages music catalog metadata, recordings, credits, agreements, and rights registrations. it enriches catalog data from global sources (ISNI, MusicBrainz, MLC, Spotify, YouTube, Discogs), reconciles identifiers across the industry identity graph, and produces registration-ready exports for collection societies and sync. pica is mcp-first. the primary experience is: your own AI client (Claude Desktop, ChatGPT, Cursor, Gemini, Grok, or any MCP-compatible tool) connects to pica and does catalog work conversationally. your AI subscription stays with the provider — pica never sits between you and your AI, never charges for AI usage, and never marks up provider pricing. writes — creating works, updating credits, submitting splits, requesting enrichment — happen through the model context protocol. there is no public REST API for writes. agents connect via the @withpica/mcp-server npm package (stdio transport) or the HTTP endpoint at https://withpica.com/api/mcp. the public directory additionally supports read-only REST at https://withpica.com/api/public/directory/* alongside its own MCP server (@withpica/mcp-server-directory). ## connecting - [connect page — browser setup](https://withpica.com/connect) - [@withpica/mcp-server](https://www.npmjs.com/package/@withpica/mcp-server) — stdio MCP server for authenticated catalog access - [@withpica/mcp-server-