SpotiClaw

Spotify Web API client for Nyx agents. Use when interacting with Spotify: search, playback, playlists, library, tracks, artists, albums, shows, podcasts. Requires SPOTIFY_CLIENT_ID, SPOTIFY_CLIENT_SECRET, SPOTIFY_REDIRECT_URI, and a local .spotify_cache token file.

Spoticlaw - Spotify Web API Client

A lightweight Spotify Web API client using direct HTTP requests. No Spotipy dependency.

Quick Start

bash

# Install dependencies
pip install requests python-dotenv

# Add to path
import sys
sys.path.insert(0, "skills/spoticlaw/scripts")

from spoticlaw import player, search, playlists, library

# Search for music
results = search().query("coldplay", types=["track"], limit=10)

# Play a track
player().play(uris=["spotify:track:..."])

# Manage playlists
playlists().create("My Playlist")
playlists().add_items("playlist_id", ["spotify:track:..."])

# Save to library
library().save(["spotify:track:..."])

Or run from the scripts directory:

bash

cd skills/spoticlaw/scripts
python -c "from spoticlaw import player; player().play(...)"

Required Configuration

Required env vars in agent runtime:

• SPOTIFY_CLIENT_ID
• SPOTIFY_CLIENT_SECRET
• SPOTIFY_REDIRECT_URI (recommended: http://127.0.0.1:8888/callback)

Required file:

• .spotify_cache (OAuth token cache)

Authentication

Security Note: Tokens never pass through the AI model. Authentication is done locally, and the token file is copied manually to the agent.

Setup

1. Create a Spotify app at https://developer.spotify.com/dashboard
2. Get CLIENT_ID and CLIENT_SECRET
3. Add http://127.0.0.1:8888/callback as Redirect URI
4. Create .env file in your LOCAL machine:

text

SPOTIFY_CLIENT_ID=your_client_id
SPOTIFY_CLIENT_SECRET=your_client_secret
SPOTIFY_REDIRECT_URI=http://127.0.0.1:8888/callback

5. Run authentication on your LOCAL machine:

bash

cd skills/spoticlaw/scripts
pip install -r requirements.txt
python auth.py

6. Open the displayed URL in your browser, authorize

7. Copy the token file to your agent:

bash

# Linux/Mac - copy to agent's skill folder
cp .spotify_cache /path/to/agent/skills/spoticlaw/.spotify_cache

# Or if agent is remote, copy via scp, USB, etc.
scp .spotify_cache user@agent:/path/to/skills/spoticlaw/.spotify_cache

That's it! No token ever touches the AI. The agent just reads the file.

Token Auto-Refresh

The library automatically handles token refresh only if the agent has the same app credentials in .env:

• Access token expires after ~1 hour
• On first API call after expiry, it uses refresh_token + client credentials to request a new access token
• Requires SPOTIFY_CLIENT_ID and SPOTIFY_CLIENT_SECRET in the agent environment
• If .spotify_cache exists but .env is missing/mismatched, refresh fails (invalid_client)
• If you get an error, run python auth.py locally again and copy updated .spotify_cache

For more on Spotify's OAuth flow, see: https://developer.spotify.com/documentation/web-api/tutorials/code-flow

Required Scopes

The auth.py script requests these scopes:

• user-read-playback-state - Read playback state
• user-modify-playback-state - Control playback
• playlist-read-private - Read private playlists
• playlist-modify-public - Modify public playlists
• playlist-modify-private - Modify private playlists
• user-library-read - Read user library
• user-library-modify - Modify user library
• user-read-recently-played - Recently played tracks
• user-top-read - Top tracks/artists
• user-follow-read - Followed artists

---

Primitives

Important: Add the module path and install deps first:

bash

pip install requests python-dotenv  # Install dependencies

# For agent execution, add to path:
import sys
sys.path.insert(0, "skills/spoticlaw/scripts")

User

python

from spoticlaw import user

user().me()  # Get current user profile

Returns: {id, display_name, email, country, ...}

---

Search

python

from spoticlaw import search

# Search for tracks
search().query("song name", types=["track"], limit=10)

# Search for artists
search().query("artist name", types=["artist"], limit=10)

# Search multiple types
search().query("coldplay", types=["track", "artist", "album"], limit=10)

Parameters:

• q: Search query string
• types: List of types: track, artist, album, playlist, show, episode, audiobook
• limit: Max 10 results (Spotify limit)
• offset: Pagination offset

---

Tracks

python

from spoticlaw import tracks

tracks().get("track_id")  # Get single track
tracks().get_multiple(["id1", "id2"])  # Get multiple tracks

Returns track metadata: {name, artists, album, duration_ms, uri, ...}

---

Artists

python

from spoticlaw import artists

artists().get("artist_id")  # Get artist details
artists().get_albums("artist_id", limit=10)  # Get artist albums

Album filters (include_groups):

• album, single, compilation, appears_on

---

Albums

python

from spoticlaw import albums

albums().get("album_id")  # Get album details
albums().get_tracks("album_id")  # Get album tracks

---

Shows (Podcasts)

python

from spoticlaw import shows

shows().get("show_id")  # Get show details
shows().get_episodes("show_id", limit=10)  # Get show episodes

---

Episodes

python

from spoticlaw import episodes

episodes().get("episode_id")  # Get episode details

---

Playlists

python

from spoticlaw import playlists, user_playlists

# Get user's playlists
user_playlists().get(limit=50)

# Get playlist details
playlists().get("playlist_id")

# Get playlist tracks
# Note: Each item has 'item' key (not 'track'), e.g. item['item']['name']
# Also can contain episodes (podcasts), not just tracks
playlists().get_items("playlist_id", limit=50)

# Create playlist
playlists().create(name="My Playlist", description="...", public=False)

# Update playlist
playlists().update("playlist_id", name="New Name")

# Add tracks
playlists().add_items("playlist_id", ["spotify:track:...", "spotify:track:..."])

# Remove tracks
playlists().remove_items("playlist_id", ["spotify:track:..."])

# Delete playlist (unfollow)
playlists().delete("playlist_id")

---

Library

python

from spoticlaw import library

library().save(["spotify:track:..."])  # Save to library
library().remove(["spotify:track:..."])  # Remove from library
library().check(["spotify:track:..."])  # Check if saved, returns [True, False]

---

Player

python

from spoticlaw import player

# Get playback state
player().get_playback_state()  # Returns {} if nothing playing
player().get_currently_playing()

# Get devices
player().get_devices()  # Returns list of devices with IDs

# Transfer playback to device
player().transfer("device_id", play=True)

# Playback control
player().play(uris=["spotify:track:..."])  # Start playing
player().play(context_uri="spotify:album:...")  # Play album/playlist
player().pause()
player().next()
player().previous()
player().seek(60000)  # Seek to 1:00 (milliseconds)
player().set_volume(80)  # Set volume 0-100

# Queue
player().add_to_queue("spotify:track:...")
player().get_queue()

# Modes
player().set_shuffle(True)
player().set_repeat("off")  # off, track, context

# Recently played
player().get_recently_played(limit=50)

---

Personalisation

python

from spoticlaw import personalisation

personalisation().get_top("tracks", time_range="medium_term", limit=20)
personalisation().get_top("artists", time_range="long_term", limit=20)

# time_range: short_term (4 weeks), medium_term (6 months), long_term (years)

---

Follow

python

from spoticlaw import follow

follow().get_followed(limit=50)  # Get followed artists

---

Composite Workflows

The primitives can be mixed and matched to create powerful automations. Here are practical examples:

Workflow 1: Play a Specific Song

python

from spoticlaw import search, player

# 1. Search for the song
results = search().query("stairway to heaven", types=["track"], limit=5)
song = results["tracks"]["items"][0]
song_uri = song["uri"]
print(f"Playing: {song['name']} by {song['artists'][0]['name']}")

# 2. Play it
player().play(uris=[song_uri])

Workflow 2: Create Playlist from Search Results

python

from spoticlaw import search, playlists

# 1. Search for songs
results = search().query("led zeppelin", types=["track"], limit=10)
track_uris = [t["uri"] for t in results["tracks"]["items"][:5]]

# 2. Create playlist
pl = playlists().create("Led Zeppelin Mix", public=False)
playlist_id = pl["id"]

# 3. Add tracks
playlists().add_items(playlist_id, track_uris)
print(f"Created playlist: {pl['name']}")

Workflow 3: Save Album to Library

python

from spoticlaw import artists, albums, library

# 1. Find artist
artist = search().query("the weeknd", types=["artist"], limit=1)["artists"]["items"][0]

# 2. Get albums
albums_list = artists().get_albums(artist["id"], include_groups="album", limit=5)

# 3. Save first album
album = albums_list["items"][0]
library().save([album["uri"]])
print(f"Saved album: {album['name']}")

Workflow 4: Play Podcast Episode

python

from spoticlaw import search, shows, player

# 1. Find podcast
podcast = search().query("joe rogan", types=["show"], limit=1)["shows"]["items"][0]
show_id = podcast["id"]

# 2. Get latest episode
episodes = shows().get_episodes(show_id, limit=1)
episode = episodes["items"][0]
episode_uri = episode["uri"]

# 3. Get device and play
devices = player().get_devices()
if devices.get("devices"):
    device_id = devices["devices"][0]["id"]
    player().transfer(device_id, play=True)
    player().play(uris=[episode_uri])
    print(f"Playing: {episode['name']}")

Workflow 5: Transfer Playback and Play

python

from spoticlaw import player, search

# 1. Get available devices
devices = player().get_devices()
print("Available devices:", [d["name"] for d in devices.get("devices", [])])

# 2. Transfer to phone
if devices.get("devices"):
    device_id = devices["devices"][0]["id"]
    player().transfer(device_id, play=True)
    
    # 3. Search and play
    results = search().query("dream on", types=["track"], limit=1)
    track_uri = results["tracks"]["items"][0]["uri"]
    player().play(uris=[track_uri])

Workflow 6: Get User's Top Artists and Follow One

python

from spoticlaw import personalisation, search, library

# 1. Get top artists
top = personalisation().get_top("artists", limit=10)
print("Your top artists:")
for i, a in enumerate(top["items"], 1):
    print(f"  {i}. {a['name']}")

# 2. Search for a new artist
new_artist = search().query("tame impala", types=["artist"], limit=1)["artists"]["items"][0]
print(f"\nFound: {new_artist['name']}")

# Note: Follow functionality requires additional scope (not in current scope list)
# Use Spotify app to follow manually

Workflow 7: Build Queue from Album

python

from spoticlaw import albums, player

# 1. Get album tracks
album_id = "4aawyAB9vmqN3uQ7FjRGTy"  # Example album
tracks = albums().get_tracks(album_id)

# 2. Add all tracks to queue
for track in tracks["items"][:5]:  # First 5 tracks
    player().add_to_queue(track["uri"])

print("Added 5 tracks to queue")

Workflow 8: Check Library for Multiple Tracks

python

from spoticlaw import library, search

# 1. Search for tracks
results = search().query("classic rock", types=["track"], limit=20)
track_uris = [t["uri"] for t in results["tracks"]["items"]]

# 2. Check which are in library
saved = library().check(track_uris)

# 3. Show results
for i, (track, is_saved) in enumerate(zip(results["tracks"]["items"], saved)):
    status = "✓ saved" if is_saved else "○ not saved"
    print(f"{i+1}. {track['name']} - {status}")

Workflow 9: Get Recently Played and Save One

python

from spoticlaw import player, library

# 1. Get recently played
recent = player().get_recently_played(limit=10)

print("Recently played:")
for i, item in enumerate(recent["items"], 1):
    track = item["track"]
    print(f"  {i}. {track['name']} - {track['artists'][0]['name']}")

# 2. Save the first one
if recent["items"]:
    track_uri = recent["items"][0]["track"]["uri"]
    library().save([track_uri])
    print(f"\nSaved: {recent['items'][0]['track']['name']}")

Workflow 10: Play from Playlist

python

from spoticlaw import playlists, player

# 1. Get user's playlists
my_playlists = user_playlists().get(limit=10)

print("Your playlists:")
for p in my_playlists["items"]:
    print(f"  - {p['name']} ({p['tracks']['total']} tracks)")

# 2. Pick first playlist and play
if my_playlists["items"]:
    playlist_id = my_playlists["items"][0]["id"]
    # Get device first
    devices = player().get_devices()
    if devices.get("devices"):
        player().transfer(devices["devices"][0]["id"], play=True)
        
        # Play playlist
        player().play(context_uri=f"spotify:playlist:{playlist_id}")
        print(f"Playing: {my_playlists['items'][0]['name']}")

---

Error Handling

python

from spoticlaw import player, SpotifyException

try:
    player().play(uris=["spotify:track:..."])
except SpotifyException as e:
    if "NO_ACTIVE_DEVICE" in str(e):
        print("No device found. Open Spotify and try again.")
    elif "Invalid token" in str(e):
        print("Token expired. Re-authenticate: python auth.py")
    else:
        print(f"Error: {e}")

---

Common Issues

Error	Solution
No token. Re-authenticate.	Run python auth.py
The access token expired	Should auto-refresh. If not, run python auth.py
Insufficient client scope	Re-auth with more scopes
NO_ACTIVE_DEVICE	Open Spotify app, then retry
Invalid limit	Use max 10 for search, 50 for playlists
Resource not found	Invalid ID or item unavailable

---

API Limits

• Search: max 10 results
• Playlist items: max 50
• Pagination: Use offset parameter
• Player: Requires active Spotify session

---

Files

• scripts/spoticlaw.py - Main API client
• scripts/auth.py - Authentication helper
• scripts/requirements.txt - Dependencies

---

More Info

For setup, troubleshooting, and contributions: https://github.com/your-org/spoticlaw