Spotify Bulk Actions Mcp

io.github.khglynn由 GitHub 帳號 khglynn 發布v0.1.1更新於 Nov 10, 2026

MCP server for bulk Spotify operations - batch playlist creation, exports, management

發布方自己的服務:actions.world

已驗證STDIO僅桌面資料與分析媒體與設計

概覽

AI 產生的概覽

本機 MCP 伺服器,用於批次處理 Spotify:從歌曲清單建立播放清單、整理播放清單,以及匯出與分析你的音樂庫。

可以做什麼
  • 讀取音樂庫與收聽資料:追蹤的藝人、收藏歌曲、常聽藝人與曲目、最近播放、音樂庫摘要
  • 搜尋曲目,包含模糊搜尋與附高/中/低信心分數的批次搜尋
  • 建立播放清單、加入曲目,並把 CSV 歌曲清單匯入成新的播放清單
  • 維護播放清單:比較、找出並移除重複曲目、批次移除、重新排序、更新資訊、匯出成 CSV
  • 批次追蹤或取消追蹤藝人,批次收藏或取消收藏曲目
  • 驗證歌曲 CSV,並匯出信心不足的比對結果供人工複核
適用情境
適合讓助理把一長串歌曲(例如播客節目筆記中的歌單)整理成 Spotify 播放清單、整理或比較既有播放清單,或統計你收藏的音樂庫與收聽紀錄。
執行需求
以本機 Python 套件形式執行(需 Python 3.10+),可從 PyPI 或程式碼倉庫安裝。需要 Spotify 帳號與 Spotify 開發者應用程式,在本地 .env 檔設定 SPOTIFY_CLIENT_ID 與 SPOTIFY_CLIENT_SECRET,並完成一次瀏覽器登入,權杖會存在本機快取目錄。需要連線至 Spotify API 的網路。
安裝前請注意
需要 Spotify 開發者憑證(SPOTIFY_CLIENT_ID、SPOTIFY_CLIENT_SECRET),並在本機保存登入權杖;請勿外流 .env 與權杖快取,若外洩應在 Spotify 後台輪替。許多工具會變更你的帳號:建立與編輯播放清單、批次追蹤/取消追蹤、收藏/取消收藏以及移除曲目。你的開發者應用程式受 Spotify 開發模式限制。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Spotify Bulk Actions Mcp,將其新增到工作區。
  2. 為需要使用其工具的對話啟用該服務。

Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。

其他 MCP 客戶端

參照 儲存庫 中的啟動說明。

README

[Spotify Bulk Actions MCP]

Spotify Bulk Actions MCP

A Model Context Protocol (MCP) server for bulk Spotify operations - batch playlist creation, library exports, and large-scale library management.

What makes this different from other Spotify MCPs?

  • Confidence scoring - Batch searches return HIGH/MEDIUM/LOW confidence for each match
  • Human-in-the-loop - Uncertain matches are exported for review, then re-imported
  • Bulk operations - Handle 500+ songs efficiently with rate limiting built-in
  • Library exports - Export your complete library data
  • Podcast playlist focused - Built specifically for importing song lists from podcast show notes

Support This Project

Made cause I can't not have headphones on, support my 80k+ pocast subscriptions. [Buy Me A Coffee]


Listed On


Projects I've Built With This

ProjectDescriptionLinks
recordOSWhich albums do you love most? A visual album collection appLive · Repo
Festival NavigatorNavigate multi-day festivals with friendsRepo

Playlists Maintained With This MCP

Coming soon: Switched On Pop, This American Life, and more podcast playlists


What This Does

Library Analysis:

  • Get all your followed artists
  • Get all saved/liked songs (handles libraries up to 10k songs)
  • Find unique artists from your library ranked by song count
  • Find albums where you have 6+ saved songs (great for vinyl shopping!)
  • Export your complete library summary

Listening Insights:

  • Your top artists and tracks (short/medium/long term)
  • Recently played history

Bulk Playlist Creation:

  • Import song lists from CSV files (for podcast playlists, etc.)
  • Batch search with confidence scoring (HIGH/MEDIUM/LOW)
  • Automatic handling of uncertain matches for human review
  • Create playlists from search results

Playlist Maintenance:

  • Find and remove duplicate tracks
  • Compare two playlists (shared vs unique tracks)
  • Bulk remove and reorder tracks, update playlist details
  • Export any playlist to CSV

Bulk Library Actions:

  • Follow/unfollow artists in bulk
  • Save/unsave tracks in bulk

Quick Start

1. Prerequisites

  • Python 3.10+
  • A Spotify account
  • Spotify Developer credentials (get them here)

2. Clone & Setup

bash
# Clone the repogit clone https://github.com/khglynn/spotify-bulk-actions-mcp.gitcd spotify-bulk-actions-mcp
# Create and activate virtual environmentpython -m venv venvsource venv/bin/activate  # On Windows: venv\Scripts\activate
# Install the packagepip install -e .
# Copy env example and add your credentialscp .env.example .env# Edit .env with your SPOTIFY_CLIENT_ID and SPOTIFY_CLIENT_SECRET

Also on PyPI: pip install spotify-bulk-actions-mcp - but you'll still need local .env and auth setup.

3. Authenticate with Spotify (One-Time)

This opens a browser for you to log in:

bash
python setup_auth.py

After login, your token is saved locally in .spotify_cache/.

4. Test It Works

bash
source venv/bin/activatepython -c "from spotify_bulk_actions_mcp.utils.auth import is_authenticated; print('Auth OK!' if is_authenticated() else 'Not authenticated')"

Option B: Hosted (claude.ai custom connector)

This server also runs as a remote MCP over streamable HTTP (MCP_TRANSPORT=http), fronted by OAuth 2.1 (WorkOS AuthKit) with a fail-closed user allowlist. That's how the author runs it: Cloud Run + Secret Manager for the Spotify refresh token, added to claude.ai as a custom connector. Tools carry readOnlyHint/destructiveHint annotations so clients can group read vs write actions. See Dockerfile and the deploy notes in the parent collection repo.

5. Connect to Claude Code

Add this to your Claude Code settings (~/.claude/settings.local.json):

json
{  "mcpServers": {    "spotify": {      "command": "/path/to/spotify-bulk-actions-mcp/venv/bin/spotify-bulk-actions-mcp"    }  }}

Restart Claude Code after adding this.

Available Tools (33)

Library Analysis

ToolDescription
check_auth_statusVerify Spotify auth is working
get_followed_artistsGet all artists you follow
get_saved_tracksGet all your liked songs
get_library_artistsArtists from saved songs, ranked by count
get_albums_by_song_countAlbums with N+ saved songs
export_library_summaryComplete library export

Listening Insights

ToolDescription
get_top_artistsYour top artists (short/medium/long term)
get_top_tracksYour top tracks (short/medium/long term)
get_recently_playedRecently played tracks

Bulk Library Actions

ToolDescription
follow_artistsFollow artists in bulk
unfollow_artistsUnfollow artists in bulk
save_tracksLike/save tracks in bulk
unsave_tracksUn-like tracks in bulk

Search

ToolDescription
search_trackSearch for a single track
search_track_fuzzyBroader search when exact fails
batch_search_tracksSearch many tracks with confidence scores
get_track_preview_urlGet 30-second preview URL

Playlist Creation

ToolDescription
create_playlistCreate a new playlist
add_tracks_to_playlistAdd tracks to existing playlist
import_and_create_playlistFull CSV → playlist workflow
create_playlist_from_search_resultsCreate from batch search
add_reviewed_tracksAdd reviewed/corrected tracks

Playlist Management

ToolDescription
get_playlist_infoGet playlist details
update_playlistUpdate name, description, public status
get_playlist_tracksGet all tracks in a playlist
export_playlist_to_csvExport a playlist to CSV
compare_playlistsShared vs unique tracks between two playlists
find_duplicate_tracksFind duplicates in a playlist
remove_duplicate_tracksRemove duplicates, keeping first occurrence
remove_tracks_from_playlistBulk remove specific tracks
reorder_playlist_tracksMove tracks within a playlist

Utilities

ToolDescription
parse_song_list_csvValidate a song CSV
export_review_csvExport uncertain matches for review

How This Differs From the Official Spotify Connector

Claude has an official Spotify connector (built by Spotify, April 2026). It's excellent at what it does — but its tool surface is small, and this MCP exists for everything it doesn't cover. As of August 2026 the official connector exposes 8 tools: Search Spotify, Create Playlist, Get Currently Playing, Add to Library, Remove from Library, Fetch Playlist Tracks, and two auth utilities.

CapabilityOfficial connectorThis MCP
Search, playback context, mood playlists✅Search only
Create a playlist✅✅ (plus CSV import w/ confidence scoring)
Add/remove Liked Songs✅ (one at a time)✅ bulk
Read your full library (saved tracks, followed artists)❌✅
Listening insights (top artists/tracks, recently played)❌✅
Edit existing playlists (bulk remove, reorder, update details)❌✅
Dedupe / compare / export playlists to CSV❌✅
Bulk follow/unfollow artists❌✅

One more difference: the official connector runs on Spotify's partner-gated MCP gateway (no dynamic client registration; tokens from ordinary developer apps are rejected), so it's only usable inside partnered AI surfaces. This MCP runs under your own Spotify developer app — Development Mode limits apply (5 allowlisted users per Client ID as of February 2026), but you control it end to end.

Example Workflows

Get Your Library Stats

Ask Claude:

"What artists do I have the most saved songs from?"

Claude will use get_library_artists and show you.

Find Albums for Vinyl

Ask Claude:

"Find albums where I have 6 or more saved songs"

Claude will use get_albums_by_song_count with min_songs=6.

Create Playlist from Song List

  1. Create a CSV file:
csv
title,artistBohemian Rhapsody,QueenHotel California,EaglesBillie Jean,Michael Jackson
  1. Ask Claude:

"Create a playlist called 'My Mix' from this CSV: [paste CSV]"

Claude will:

  1. Parse the CSV
  2. Search each song with confidence scoring
  3. Create the playlist with high-confidence matches
  4. Show you uncertain matches to review

Bulk Podcast Playlist

For large lists (500+ songs):

  1. Ask Claude to use batch_search_tracks with your song list
  2. Review the results (HIGH goes in automatically)
  3. Use export_review_csv to get uncertain matches
  4. Review/correct in a spreadsheet
  5. Use add_reviewed_tracks to add your corrections

Rate Limits

The server handles Spotify's rate limits automatically:

  • Small delays between API calls
  • Automatic retry on 429 errors
  • Caching to reduce repeat calls

For 10k songs, expect the initial library fetch to take 2-3 minutes.

Files & Data

LocationPurpose
.envYour Spotify credentials (gitignored)
.spotify_cache/Auth tokens and cached data (gitignored)
spotify_bulk_actions_mcp/server.pyMain MCP server
spotify_bulk_actions_mcp/tools/Tool implementations

Troubleshooting

"Not authenticated" error:

bash
python setup_auth.py

Rate limit errors: Wait a few minutes and try again. The server will auto-retry.

Token expired: The server auto-refreshes tokens. If issues persist, re-run setup_auth.py.

Security Notes

  • Your credentials are in .env (gitignored, never committed)
  • Auth tokens are stored locally in .spotify_cache/
  • Never share your .env or token files
  • If credentials are exposed, rotate them in Spotify Dashboard

License

MIT


Made cause I can't not have headphones on. If this helps you, buy me a coffee!

來源:README.md,提交 97dba94

工具

0
SourceWeft 尚未在沙箱中啟動過該軟體包,因此其工具未知。

版本歷史

1
  1. v0.1.1最新Sep 16, 2026

評分

已滿足適用檢查項的 50%

該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通25 分

    請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。

  • 已宣告認證方式10 分

    請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。

  • 已列出工具15 分

    請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。

  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

其他 Actions 整合