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 skills/labelgrid/labelgrid-mcp/labelgrid-releasenpx skills add labelgrid/labelgrid-mcp --skill labelgrid-releasegit clone --depth 1 https://github.com/labelgrid/labelgrid-mcpWrote 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/skills/labelgrid/labelgrid-mcp/labelgrid-release)<a href="https://agentmods.dev/skills/labelgrid/labelgrid-mcp/labelgrid-release"><img src="https://agentmods.dev/badge/skills/labelgrid/labelgrid-mcp/labelgrid-release.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 | $0.00061 | $0.03306 |
| Opus 5 | $0.00030 | $0.01653 |
| Sonnet 5 | $0.00012 | $0.00661 |
| Haiku 4.5 | $0.00006 | $0.00331 |
Grade A, and why
labelgrid-release 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 — 254 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Releasing music through LabelGrid
This skill walks an agent through the full release lifecycle on the LabelGrid public API, and — just as importantly — tells you which vehicle to reach for at each step.
LabelGrid ships two tools over the same API surface (both thin wrappers — every rule and validation lives on the server, so the two behave identically):
@labelgrid/mcp— the Model Context Protocol server. 30 consolidated tools your AI client calls directly. Best for reasoning steps: drafting metadata, reading validation output, inspecting review issues.@labelgrid/cli— thelabelgridcommand-line tool. The same API from a shell, with--jsonfor pipelines. Best for moving file bytes and for batch/CI work.
Everything below names the exact tool, selector, flag, and environment variable for each vehicle. Nothing here is invented — if a step exists in only one vehicle, that is called out.
1. Pick the right vehicle
| Step type | Vehicle | Why |
|---|---|---|
| Judgment / reasoning — draft release & track metadata, resolve genre/language/outlet IDs, read validation errors, inspect review issues | MCP tools | The agent reasons over structured JSON already in context; no shell required. |
| Byte-moving & batch — upload audio/artwork, download masters/statements, loop over many releases or tracks, run in CI | CLI (labelgrid …) |
Streams file bytes, pipes --json to jq, and drives shell loops. Available only when the environment can execute shell commands. |
| Final gated actions — distribute, takedown, confirm a QC hold | Either | The MCP path is gated behind the full-writes acknowledgment (env vars); the CLI path prompts for interactive confirmation. |
Rule of thumb: reason with MCP, move files and batch with the CLI. When only one is available, it can do the whole job — each vehicle gates the consequential steps its own way (full-writes arming for MCP, confirmation prompts for the CLI); neither path skips the gate.
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 · 254 lines · 61 tokens per session scan A 3216b12a887f
labelgrid-release is a skill published in the GitHub repository labelgrid/labelgrid-mcp (3 stars, last pushed 2d ago), licensed MIT. It adds 61 tokens to every session and 3,306 once invoked, about $0.0003 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 skills, from other repositories
pr
Use for any GitHub Pull Request work. Auto-detects lifecycle phase (create / update / report) from current state — only requires explicit command for status, comments, or merge. Triggered by /pr, 'PR', 'create PR', 'open PR', 'update PR', 'PR status', 'merge PR'. Suggests, never auto-creates without user invocation.
scheduler
Use when work should happen later or repeatedly — 'schedule', 'routine', 'nightly', 'every morning', 'cron', 'remind me later', 'check back in an hour', 'watch this PR', '/scheduler'. Picks the right scheduler for the lifetime the work needs (cloud routine / session loop / desktop task), creates, lists, updates and…
ci
Use when the user wants CI status, failed-job logs, or help fixing a red build. Triggered by /ci, 'CI status', 'check the build', 'fix CI', 'why is CI failing', 'look at the build'. Auto-routes by state: status / logs / fix-proposal. Reads logs locally — never re-triggers builds without explicit user command.
orca
Orchestrator mode — the default working mode of this project, not a toggle to find. The main agent does no task work itself: every unit goes to a role-framed subagent that inherits its model and effort, at most 5 in parallel unless overridden. Load it for the contract (roles, width, write scopes, verification), when…
done
Use when the user signals work completion with 'done', 'fertig', 'finished', 'abschluss', '/done'. Detects current branch (main vs feature), runs closure checks defined in CLAUDE.md, handles commit and push based on branch context, closes related GitHub issues, and reports a strict short summary.
gitnexus-cli
Use for read-only GitNexus index maintenance: check index status, list indexed repos, register or (re)build the index. Examples: "Is the GitNexus index current?", "List indexed repos", "Register this repo".