Skip to main content
Glama

Last.fm MCP Server

A Model Context Protocol (MCP) server for Last.fm. Gives AI assistants access to your listening history, music discovery, and detailed track/artist/album information.

Runs on Cloudflare Workers. Your MCP client signs you in with OAuth the first time it connects, and after that every tool is available, from track and artist lookups to your own listening history.

Quick start

lastfm-mcp.com has setup steps for each app, including ChatGPT, Codex, Cursor, Devin Desktop (formerly Windsurf) and OpenCode. If an AI agent is setting it up for you, point it at lastfm-mcp.com/llms.txt.

Claude.ai / Claude Desktop

  1. Go to Customize -> Connectors

  2. Click Add, then Add custom connector. Name it Last.fm, enter https://lastfm-mcp.com/mcp and click Continue

  3. Leave the sign-in settings as they are, click Add, then Connect, and sign in to Last.fm

Claude Code

claude mcp add --scope user --transport http lastfm "https://lastfm-mcp.com/mcp"

Then start Claude Code, run /mcp, choose lastfm, then Authenticate, and sign in to Last.fm in your browser.

Other MCP clients

For clients that don't support remote servers directly (Continue.dev, Zed, etc.), use mcp-remote:

{
	"mcpServers": {
		"lastfm": {
			"command": "npx",
			"args": ["-y", "mcp-remote", "https://lastfm-mcp.com/mcp"]
		}
	}
}

MCP Inspector

npx @modelcontextprotocol/inspector https://lastfm-mcp.com/mcp

Related MCP server: lastfm-mcp

Authentication

The server uses OAuth 2.0. The first time your client connects, your browser opens a short page on lastfm-mcp.com that names the app asking for access and the address that access will go to. Approve it, then sign in to Last.fm. That browser remembers the approval for a year, so reconnecting the same app skips the page.

Access is read-only: the server can read your listening history and profile, but it can't scrobble or change anything. Sign-ins expire after 30 days, and your client then asks you to sign in again.

Available tools

Lookups and diagnostics

Tool

Description

get_track_info

Detailed information about a track

get_artist_info

Artist information and bio

get_album_info

Album details and track listing

get_artist_top_tracks

An artist's globally most-played tracks

get_artist_top_albums

An artist's globally most-played albums

get_similar_artists

Artists similar to a given artist

get_similar_tracks

Tracks similar to a given track

ping

Test connectivity

server_info

Server status and capabilities

lastfm_auth_status

Check authentication status

Your listening data

Tool

Description

get_recent_tracks

Recent listening history (paginated)

get_top_artists

Top artists by time period

get_top_albums

Top albums by time period

get_top_tracks

Top tracks by time period

get_loved_tracks

Loved/favorited tracks

get_user_info

Last.fm profile information

get_listening_stats

Listening statistics

get_music_recommendations

New-to-you artists, with the reason for each pick and starter tracks

Your listening history over time

Tool

Description

get_weekly_chart_list

Available historical time periods

get_weekly_artist_chart

Artist charts for a specific time period

get_weekly_track_chart

Track charts for a specific time period

Good for questions like "when did I start listening to Led Zeppelin?" or "what was I into last March?"

Resources

The server exposes MCP resource URIs:

lastfm://user/{username}/recent          # Recent tracks
lastfm://user/{username}/top-artists     # Top artists
lastfm://user/{username}/top-albums      # Top albums
lastfm://user/{username}/loved           # Loved tracks
lastfm://user/{username}/profile         # User profile
lastfm://track/{artist}/{track}          # Track info
lastfm://artist/{artist}                 # Artist info
lastfm://album/{artist}/{album}          # Album info
lastfm://artist/{artist}/similar         # Similar artists
lastfm://track/{artist}/{track}/similar  # Similar tracks

Prompts

Prompt

Description

Arguments

listening_insights

Analyze listening habits and patterns

username, period?

music_discovery

Discover music based on listening history

genre?

track_analysis

Detailed analysis of a track

artist, track

album_analysis

Detailed analysis of an album

artist, album

artist_analysis

Detailed analysis of an artist

artist

listening_habits

Summarize listening habits

username, timeframe?

Development

Prerequisites

  • Node.js 22 (see .nvmrc; CI runs 22)

  • Cloudflare Workers account

  • Last.fm API key (create one here)

Local setup

git clone https://github.com/rianvdm/lastfm-mcp.git
cd lastfm-mcp
npm install

Create .dev.vars (or copy .dev.vars.example):

LASTFM_API_KEY=your_api_key
LASTFM_SHARED_SECRET=your_shared_secret
CONSENT_SECRET=any_random_string_of_32_or_more_characters
npm run dev

Test with the inspector:

npx @modelcontextprotocol/inspector http://localhost:8787/mcp

Deployment

Set production secrets (or run npm run setup:prod, which creates the KV namespaces, prompts for the two Last.fm values and generates CONSENT_SECRET):

echo "your_api_key" | wrangler secret put LASTFM_API_KEY --env production
echo "your_shared_secret" | wrangler secret put LASTFM_SHARED_SECRET --env production
openssl rand -base64 32 | wrangler secret put CONSENT_SECRET --env production

CONSENT_SECRET signs the cookie that remembers which apps a browser approved. Rotating it shows the consent page again to everyone; if it's missing or shorter than 32 characters, approvals aren't remembered, but sign-ins still work.

Upgrading a deployment from v2.6.0 or earlier: v2.7.0 added the consent page and the CONSENT_SECRET secret. Set it with the openssl line above before you deploy, or right after. Until it's set, the consent page shows on every sign-in.

Deploy:

npm run deploy:prod

Testing

npm test          # vitest, in workerd via @cloudflare/vitest-pool-workers
npm run lint
npm run build     # wrangler deploy --dry-run: catches type and binding errors

Architecture

  • Runtime: Cloudflare Workers

  • Protocol: MCP (streamable HTTP)

  • Auth: OAuth 2.0 (RFC 9728) via @cloudflare/workers-oauth-provider; the provider owns sessions and tokens

  • Storage: Cloudflare KV for OAuth state, sessions, caching, and rate-limit counters

  • Rate limiting: per-IP on /mcp, so one client can't exhaust the shared Last.fm API key's budget

  • API: Last.fm Web API v2.0

Endpoints

Endpoint

Purpose

/mcp

MCP JSON-RPC endpoint (rate-limited per IP)

/authorize

OAuth 2.0 authorization

/oauth/token

OAuth 2.0 token exchange

/oauth/register

OAuth 2.0 dynamic client registration

/.well-known/oauth-authorization-server

OAuth server metadata

/.well-known/oauth-protected-resource

OAuth resource metadata

/.well-known/mcp.json

MCP server card (name, version, transport)

/llms.txt

Setup instructions for AI agents

/health

Health check

/

Landing page

Contributing

  1. Fork the repo

  2. Create a feature branch

  3. Commit your changes

  4. Open a pull request

License

MIT

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      B
      maintenance
      Search, rate and organise your Discogs collection, wantlist and folders from an AI assistant. Self-hosted: deploy your own copy to Cloudflare Workers with the Deploy to Cloudflare button in the README. Not affiliated with Discogs.
      20
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      An MCP server for the Last.fm API that enables LLM agents to access music data including user profiles, listening history, top charts, search, and artist/album/track metadata.
      16
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      An MCP server for the Last.fm API, exposing artist, album, track, chart, and user data as read-only tools. Supports optional writes, auth, and experimental methods over stdio or HTTP.
      MIT