Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/kud/mcp-qobuz/copilot-instructionsgit clone --depth 1 https://github.com/kud/mcp-qobuzWrote 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.
[](https://agentmods.dev/instructions/kud/mcp-qobuz/copilot-instructions)<a href="https://agentmods.dev/instructions/kud/mcp-qobuz/copilot-instructions"><img src="https://agentmods.dev/badge/instructions/kud/mcp-qobuz/copilot-instructions.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01416 | $0.01416 |
| Opus 5 | $0.00708 | $0.00708 |
| Sonnet 5 | $0.00283 | $0.00283 |
| Haiku 4.5 | $0.00142 | $0.00142 |
Grade A, and why
mcp-qobuz copilot-instructions.md 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Copilot Instructions — mcp-qobuz
Project context
@kud/mcp-qobuz is a TypeScript MCP server that exposes the Qobuz music service
to AI assistants via the Model Context Protocol. It is one of four surfaces in the
@kud/qobuz ecosystem; all share the @kud/qobuz core client library.
Stack:
- Runtime: Node.js >=20, ESM-only (
"type": "module", NodeNext resolution) - MCP layer:
@modelcontextprotocol/sdk— tools registered withserver.registerTool() - Input validation: Zod v4 schemas passed directly to
registerToolinputSchema - Core client:
@kud/qobuz(auth, API transport, types, deep-link helpers) - Auth: macOS Keychain (
createKeychainStore) or env vars (QOBUZ_TOKEN+QOBUZ_APP_ID) - Tests: Vitest — unit tests mock
@kud/qobuzviavi.fn(), exportmakeHandlersfor testability - Build:
tsc→dist/, strict mode, no bundler
Key architectural invariants:
- All tool handlers live inside
makeHandlers(client)and returnok(data)orerr(message)— never throw out of a handler. - Write tools (
create-playlist,add-to-playlist,update-playlist-description) requireconfirm: trueto execute; they return anerr()message otherwise. now-playingis macOS-only and reads a local Qobuz desktop app state file; it must return a gracefulerr()on any other platform or when the file is absent.console.erroris the correct logging channel (stdout is the MCP stdio transport).- The
convert-linktool is intentionally absent pending extraction of a shared@kud/song-matchpackage. Do not copy the resolve logic into this repo.
Content rules
Flag these patterns:
-
Uncaught throws in tool handlers — every handler must catch its own errors and return
err(msg). A throw propagating out of a handler crashes the MCP connection rather than returning a structured error to the client. -
console.log in server code — only
console.erroris safe;console.logwrites to stdout which is the MCP stdio transport and will corrupt the protocol stream.
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.
- 4d ago First seen · 133 lines · 1,416 tokens per session scan A f7c55dd8926d
mcp-qobuz copilot-instructions.md is an instructions file published in the GitHub repository kud/mcp-qobuz (0 stars, last pushed 24d ago), licensed MIT. It adds 1,416 tokens to every session, about $0.0071 per session on Opus 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 2026-08-31.
Other instructions, from other repositories
mcp-jenkins CLAUDE.md
Instructions for kud/mcp-jenkins, covering mcp-jenkins, api reference, distribution and public surface.
fcp-midi CLAUDE.md
Instructions for os-tack/fcp-midi, covering fcp-midi, project overview, architecture, key directories and commands.
roon-mcp AGENTS.md
AGENTS.md instructions for bestimmaa/roon-mcp, covering roon mcp — project notes and references.
soundiiz-mcp AGENTS.md
AGENTS.md instructions for BASIC-BIT/soundiiz-mcp, covering agents, workflow, tool ergonomics goals (distilled), curated output philosophy and specific notes (current direction).
soundiiz-mcp CLAUDE.md
Claude Code instructions for BASIC-BIT/soundiiz-mcp: The agent guidance for this repo lives in AGENTS.md. That file is the single source of truth — Claude Code, OpenCode, and any other AGENTS.md-aware tool all read from there.
mcp-lastfm CLAUDE.md
Claude Code instructions for kud/mcp-lastfm, covering mcp-lastfm and api reference.