sonovault-mcp
Provides access to the SonoVault music metadata API, enabling search and retrieval of information about 90M+ tracks, artists, labels, and releases, including ISRC, ISWC, and cross-platform IDs.
README
sonovault-mcp
MCP server for the SonoVault music metadata API. Gives Claude, Cursor, and any MCP client search across 90M+ tracks, artists, labels, and releases, with full metadata per result: genre, label, release dates, artist credits, ISRC and ISWC codes, and cross-platform IDs (Spotify, Apple Music, Tidal, Beatport, Discogs, MusicBrainz, YouTube).
Ask your assistant things like:
- "What releases did mau5trap put out this year?"
- "Find Daft Punk's Veridis Quo and give me its genre, label, and ISRC"
- "Show me the tracklist of Discovery with ISRCs"
- "What's the Spotify and Beatport ID for ISRC GBDUW0000053?"
- "Resolve this play log to ISRCs and labels" (paste up to 100 lines)
- "List every recording of ISWC T-071.020.439-9"
Setup
You need a SonoVault API key. The free tier is 1,000 requests/month, no credit card: get a key.
Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"sonovault": {
"command": "npx",
"args": ["-y", "sonovault-mcp"],
"env": { "SONOVAULT_API_KEY": "svk_live_..." }
}
}
}
Claude Code
claude mcp add sonovault -e SONOVAULT_API_KEY=svk_live_... -- npx -y sonovault-mcp
Cursor
Add to ~/.cursor/mcp.json (or .cursor/mcp.json in your project):
{
"mcpServers": {
"sonovault": {
"command": "npx",
"args": ["-y", "sonovault-mcp"],
"env": { "SONOVAULT_API_KEY": "svk_live_..." }
}
}
}
VS Code (GitHub Copilot)
Add to .vscode/mcp.json:
{
"servers": {
"sonovault": {
"command": "npx",
"args": ["-y", "sonovault-mcp"],
"env": { "SONOVAULT_API_KEY": "svk_live_..." }
}
}
}
Windsurf
Add the same mcpServers block as Claude Desktop to ~/.codeium/windsurf/mcp_config.json.
Other MCP clients
Any client that speaks MCP over stdio works. Command: npx -y sonovault-mcp with the SONOVAULT_API_KEY environment variable set.
Tools
| Tool | What it does |
|---|---|
search_tracks |
Fuzzy search by artist + title. Returns ISRC, genre, label, release date. |
lookup_isrc |
Exact recording lookup by ISRC. |
get_track |
Fetch a track by SonoVault ID. |
get_work_codes |
Recording to composition: the ISWC(s) behind a recording. |
get_recordings_of_work |
Composition to recordings: every recording of an ISWC. |
get_platform_links |
One track's ID and deep link on all 7 platforms, from any ID or ISRC. |
resolve_tracks |
Bulk resolve up to 100 track names, ISRCs, or platform IDs. |
search_artists, get_artist_releases |
Artist search and discography. |
search_labels, get_label_releases |
Label search and catalog. |
search_releases, get_release |
Release search and tracklists with ISRCs. |
list_genres |
The canonical genre hierarchy. |
browse_tracks |
Discover by label, artist, genre, or year (paid tier). |
All tools are read-only. Most work on the free tier; browse_tracks needs a paid plan.
Example workflows
Enrich a radio play log for royalty reporting
Paste a play log with just artist and title, and ask: "Resolve these to ISRC, album, and label, and give me a table I can map onto a SoundExchange Report of Use." The assistant calls resolve_tracks once per 100 lines and returns the identifiers collecting societies match on. Works the same for PPL, Sena, GVL, and Re:Sound reporting.
Clean up a DJ library
"Here are 50 tracks from my rekordbox export. Find the canonical version of each, flag duplicates that share an ISRC, and tag each with its genre." The assistant combines resolve_tracks with the genre fields on each result.
Bridge recordings to compositions
"I have ISRCs from my distributor. Which musical works do they belong to, and what other recordings exist of each work?" The assistant chains get_work_codes and get_recordings_of_work. Useful for publishing administration and cover/version research.
Check cross-platform availability
"Is this Beatport exclusive on Spotify yet? Beatport ID 6014288." One get_platform_links call answers it, with deep links per platform.
Research a label
"Summarize what Anjunabeats released in the last six months, grouped by subgenre." The assistant pages through get_label_releases and pulls genre from the tracklists.
FAQ
Is there a free tier?
Yes. A free SonoVault key includes 1,000 requests/month with no credit card. Every tool except browse_tracks works on it.
What metadata does each track include?
Title, artist credits (with primary/remixer flags), ISRC, duration, genre and subgenre, and per-release album title, label, and release date. ISWC work codes come from the dedicated work-code tools.
Does it return BPM or musical key?
No. The public SonoVault API deliberately excludes audio features like BPM, key, and energy. No audio previews or artwork either.
Where does the data come from?
Aggregated and deduplicated from MusicBrainz, Discogs, Spotify, Apple Music, Tidal, and Beatport, with ISWC links from The MLC. Details: sonovault.now/data-sources.
How is this different from a Spotify API integration?
Single-platform APIs return their own IDs only. SonoVault resolves one recording to its IDs on all seven supported platforms, keyed by ISRC, and adds identifiers (ISRC, ISWC) that platform APIs don't expose. Comparison: sonovault.now/compare/spotify-api.
Is my API key sent anywhere besides SonoVault?
No. The server runs locally over stdio and sends your key only to api.sonovault.now as a request header. The code is small and auditable: src/server.ts.
Agent Skill
Not using MCP? skill/SKILL.md is an Agent Skill that teaches any agent to call the SonoVault API directly over HTTP: search rules, ISRC/ISWC recipes, bulk resolution, and error handling.
Related
- SonoVault API docs
- sonovault-js and sonovault-python, the client SDKs (this server is built on sonovault-js)
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。