music-flow

music-flow is a skill for Claude Code from P47Phoenix/Claude-Plugins. It costs 120 tokens per session (5,505 once invoked), scanned A, original, MIT.

A staged songwriting workflow that coordinates a team of specialized agents from the initial brief through concept, structure, lyrics, music, arrangement, and release. It uses checks and saved files so a run can resume later.

In plain words
What is it for?
It detects the song type, dispatches the appropriate creative roles, manages review loops and challenge checks, tracks decisions, and coordinates final release.
Why use it?
It keeps song creation organized and separates writing, judging, and approval. Automated checks decide pass or fail, while human overrides are recorded.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: model in frontmatter; mentions subagents.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the music-team plugin — 9 skills shipped together

Good fit It detects the song type, dispatches the appropriate creative roles, manages review loops and challenge checks, tracks decisions, and coordinates final release.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add P47Phoenix/Claude-Plugins
Claude Code
/plugin install music-team

Made for: Claude Code.

Or install music-team, the plugin that ships this one along with the rest of its 9 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for music-flow

README.md
[![agentmods](https://agentmods.dev/badge/skills/p47phoenix/claude-plugins/music-flow/github.svg)](https://agentmods.dev/skills/p47phoenix/claude-plugins/music-flow)
Your own site
<a href="https://agentmods.dev/skills/p47phoenix/claude-plugins/music-flow"><img src="https://agentmods.dev/badge/skills/p47phoenix/claude-plugins/music-flow/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for music-flow

Your own site · 80×15
<a href="https://agentmods.dev/skills/p47phoenix/claude-plugins/music-flow"><img src="https://agentmods.dev/badge/skills/p47phoenix/claude-plugins/music-flow.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 120 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 5,505 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00120 $0.05505
Opus 5.5 $0.00048 $0.02202
Sonnet 5.5 $0.00024 $0.01101
Haiku 4.5 $0.00012 $0.00551

Measured 2d ago against content hash 039ed3da8726, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

music-flow scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

music-team/skills/music-flow/SKILL.md · 330 lines

How it starts

The opening of the file, as written. The whole thing — 330 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Music Flow Orchestrator

Design Principle

This skill is the ORCHESTRATOR. It coordinates the song team through a staged pipeline and NEVER produces creative content: no title, hook, lyric line, chord cell, ABC note, or texture note. Every creative artifact is written by a named sub-agent (lyricist, composer, arranger, adapter, song-critic) dispatched with the Agent tool. Full doctrine, anti-patterns, and guardrails: music-team/references/shared/orchestrator-doctrine.md.

Invariants (one line each, rationale in the doctrine):

  1. The orchestrator never writes the song; its only write paths are .song/active, .song/<slug>/state.json, .song/<slug>/decisions.md, .song/memory/**.
  2. One role = one Agent call. Never combine writer and judge, or two roles, in one prompt.
  3. Scripts decide pass or fail. Sub-agents never self-certify; only the user overrides a gate, and every override is recorded.
  4. The challenger sees final artifacts only, never drafts or critiques.
  5. Artifacts move by path. Signals stay at 200 characters or fewer.
  6. Everything is on disk; state.json is rewritten after every stage so any session can resume.

Model awareness (Opus 4.7): under-dispatch is the main regression mode. The number of Agent calls at a stage MUST equal the roles listed in references/stages.yml for that song type and cell. A short-count dispatch, or doing a role's work inline, is a violation.

Plugin root: resolve <plugin-root> once in Phase 0 (${CLAUDE_PLUGIN_ROOT} when set, else music-team/) and use literal paths afterwards. Shared references live under <plugin-root>/references/shared/; this skill's under <plugin-root>/skills/music-flow/references/.


Phase 0: Setup and Resume

Run before anything else.

  1. Config. Read .song/config.yml. If absent, music-brief creates it from <plugin-root>/.song.config.yml.example and announces it. Keys and defaults: references/shared/config-schema.md. Note quick, routing.force_type, loops.*, escalation.*, adaptation.*, memory.*, quality.*.
  2. Quick mode. When quick: true or the request says "quick song": every loops.<stage> is 1 (including release), escalation.on_loop_exhaustion is accept_best, the wizard asks at most 2 questions. Record quick: true in state.json. Quick mode never changes which stages run.
  3. Agent roster check. Confirm each agent file <plugin-root>/agents/<name>.md exists for the agents the routed stages will use. A missing file is a clear error, not a fallback: MISSING_AGENT: <name> (agents/<name>.md not found). Stop that stage; never substitute general-purpose or perform the role inline.
  4. Resume check. Read .song/active.
    • Names a slug whose .song/<slug>/state.json has status: in_progress: tell the user Resuming <slug> at stage <N> (<name>), re-run that stage's gates first, then continue from state.stage. Never start a second run while one is active; offer resume, or abort first.
    • Names a slug with status: complete and the request is a change: route as REVISE (Phase 1).
    • status: failed: show the failure, offer revise <stage> or start for a new song.
    • Absent or empty: proceed to Phase 1.
  5. Command parse. If the request is a User Command (below), execute that command and stop.

Read the full file on GitHub · 330 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 330 lines · 120 tokens per session scan A 039ed3da8726

Subscribe to this mod's changes

music-flow is a skill published in the GitHub repository P47Phoenix/Claude-Plugins (2 stars, last pushed 2d ago), licensed MIT. It adds 120 tokens to every session and 5,505 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-26.

Related

Other skills, from other repositories

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

insight-error-page

Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…

vercel/next.js · 83 tokens