From recoup-songs
Analyze a song, track, or audio file with Recoup's audio language model (Music Flamingo). Use when asked to "analyze this song", "what does this track sound like", "run Music Flamingo", "use the audio model", "get a full report on this audio", "what key/BPM/mood/genre is this", "transcribe the lyrics", or any request to understand a piece of audio from its sound. This is the default entry point for audio analysis — it resolves the audio URL, picks the right preset, calls the endpoint, and routes to focused workflows (metadata, lyrics, playlist pitch, sync brief, mix feedback) when the ask is specific.
How this skill is triggered — by the user, by Claude, or both
Slash command
/recoup-songs:recoup-song-analyzerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Turn a song's **audio** into structured intelligence — genre, key, BPM, mood,
Turn a song's audio into structured intelligence — genre, key, BPM, mood, lyrics, mix notes, sync fit, audience, and more — using Recoup's audio language model. This is the entry point for "analyze this track" requests.
The endpoint contract (auth, base URL, request/response, every preset, call
patterns, gotchas) lives in references/flamingo-api.md. Read it before calling.
Match the user's ask to the smallest useful action:
full_report only when they want everything.catalog_metadata
(add music_theory for chords/sections).recoup-song-metadata.recoup-song-lyrics.recoup-song-playlist-pitch.recoup-song-sync-brief.recoup-song-mix-feedback.Each recoup-song-* skill is self-contained — if one is installed, prefer it for
its job. If only this skill is present, run the relevant preset directly.
audio_url — it
cannot take a local file. If the user gives a URL (or a Recoup/chat attachment
URL), use it. If they only have a local file, explain the URL requirement and
ask for a hosted link rather than pretending the file can be sent.full_report. Pick
the one preset (or short chain) that answers the question. See the preset table
in references/flamingo-api.md. When unsure which preset fits, list them with
GET /api/songs/analyze/presets.analyze_music if the host exposes it; use the recoup
CLI only when RECOUP_API_KEY is set. Exact commands are in
references/flamingo-api.md.| User wants | Preset(s) |
|---|---|
| Genre / mood / BPM / key / instruments | catalog_metadata |
| Vibe tags only | mood_tags |
| Chords, scale, sections | music_theory |
| Lyrics | lyric_transcription |
| Explicit / brand-safety / radio check | content_advisory |
| Marketing blurb | song_description |
| Comparable artists | similar_artists |
| Sample/interpolation risk | sample_detection |
| Everything | full_report (13 presets in parallel — slower, more spend) |
If you are inside an artist workspace, save durable results under
songs/{song-slug}/analysis/:
music-flamingo-full-report.json — raw full_report output{preset}.json or {preset}.md — targeted preset outputanalysis-summary.md — your human-readable synthesisIf there is no workspace, just return the synthesis in the conversation.
audio_url.full_report runs 13 model calls; reach for single
presets first.account_id in the request body. Auth determines the account.references/flamingo-api.md — auth, base URL, request/response, full preset
table, REST/CLI/MCP call patterns, and gotchas.Searches MemPalace before answering questions about past work, people, projects, or prior decisions. Returns verbatim stored content instead of guessing from model memory.
Guides Payload CMS config (payload.config.ts), collections, fields, hooks, access control, APIs. Debugs validation errors, security, relationships, queries, transactions, hook behavior.
Implements vector databases with Pinecone, Weaviate, Qdrant, Milvus, pgvector for semantic search, RAG, recommendations, and similarity systems. Optimizes embeddings, indexing, and hybrid search.
npx claudepluginhub recoupable/skills --plugin recoup-skills