audio-analysis

Inspect downloaded track waves or query APIs to analyze audio parameters like BPM, Key, Energy, and Valence.

SKILL.md

GitHub

Audio & Music Analysis Skill

Overview

This skill guides AI agents in performing deep signal-processing analysis or metadata lookups on music tracks. By analyzing audio parameters such as BPM (Beats Per Minute), Musical Key / Scale (e.g. C Major, A Minor), and energy metrics, the agent can help users categorize playlists, match tempos for workouts or DJ sets, and explore composition structures.


Technical Analysis Approaches

Approach 1: Audio Metadata API Integration

Agents can query open music repositories (such as AcousticBrainz, Spotify Audio Features, or MusicBrainz) using song metadata to fetch precise pre-computed acoustic features.

{
  "title": "晴天",
  "singer": "周杰伦",
  "bpm": 84,
  "key": "G Major",
  "valence": 0.52,
  "energy": 0.48,
  "danceability": 0.58,
  "time_signature": "4/4"
}

Approach 2: Native Audio Waveform Extraction (CLI Signal Processing)

When the track is downloaded locally via alx download <id>, the agent can run local CLI signal processing tools (such as aubio, ffmpeg, or custom scripts using librosa / essentia / madmom models) to analyze the audio file.

1. BPM / Tempo Detection

Identify the rhythmic rate of the song:

# Using aubio CLI tool to detect tempo (BPM) on downloaded track
aubiopitch -i "/path/to/song.mp3"
aubiotempo -i "/path/to/song.mp3"

2. Key & Scale Detection

Analyze spectral pitch classes (chroma) to determine the tonic key:

# Conceptual ESSENTIA key extractor Python script
import essentia.standard as es
loader = es.MonoLoader(filename="song.flac")
audio = loader()
key_extractor = es.KeyExtractor()
key, scale, strength = key_extractor(audio)
print(f"Key: {key} {scale} (Strength: {strength})")

Agent Usage Patterns

Pattern 1: Automatic Tempo-Matched Playlist

Build a playlist with songs matching a target BPM range (e.g. 120-130 BPM for jogging):

  1. Query search cache or local music files.
  2. Filter tracks matching the desired tempo.
  3. Automatically load them into a running queue:
# Retrieve track metadata and filter for workout tempo (e.g. 125 BPM)
alx search "workout hits" --json | jq -r '.list[] | select(.bpm >= 120 and .bpm <= 130) | .id' | while read id; do
    alx queue add "$id"
done

Pattern 2: Harmonious Transition Analysis

Advise the user on key matches (Camelot Wheel / Circle of Fifths) for smooth playlist progression:

  • Track A (G Major / 9B) transitions harmoniously into Track B (D Major / 10B, C Major / 8B, or E Minor / 9A).