Muse Code Skills
Muse is a category of agent skills: SKILL.md packages an AI agent can load to work with Muse, the AI-powered tool. Skills in this topic cover the tasks developers typically hand to an agent around Muse — connecting to it, driving its features from an agent workflow, and folding it into a larger pipeline. Read the ranked list as a shortlist, not a catalog. Start with the entries whose descriptions match the job you actually need done, then check what each skill requires before you load it: the SKILL.md should state its dependencies and the inputs it expects. If two skills overlap, prefer the one whose scope is narrower and whose setup you can satisfy today. Load one, test it against a real task, and only then add a second.
6 to start with7 top options
Start with these.
An Agent Skill fits here when it runs inside this agent.
These are the closest matches for this work. Start with the first one and open it to install.
- Rank 01View Agent SkillDirect
muse-delegation
NotableDelegates tasks to Meta's Muse Code CLI (muse) via delegation-core. Use when delegation-core selects Muse or repository-scale work needs a 1M-token context.
- Rank 02View Agent SkillDirect
ralph
NotableRun a PRD-driven implementation loop against .omm/state/prd.json, iterating story by story until every acceptance criterion passes, then a final reviewer verification pass. Use ONLY when explicitly invoked (/ralph) against an approved PRD, normally right after /ralplan. Do NOT use without an approved prd.json (run /ralplan first), and do not invoke this for a single small edit — the loop overhead only pays for itself across multiple testable stories.
- Rank 03View Agent SkillDirect
ralplan
NotableTurn an approved spec into a consensus-reviewed plan with a PRD (prd.json) of testable stories, gating vague /ralph or /team requests before they run unattended. Use ONLY when explicitly invoked (/ralplan), typically right after deep-interview hands off a spec, or when the user asks to plan before a ralph/team run. Do NOT use for direct small edits, and do not treat an invocation of /ralph or /team as implicitly requesting this — ralplan only runs when named.
- Rank 04View Agent SkillDirect
cancel
NotableEnd any active pipeline stage (deep-interview, deep-dive, trace, ralplan, ralph, team) and clean up .omm/ state left running or in progress. Use ONLY when the user explicitly asks to cancel, stop, or abandon current work, or invokes /cancel. Do NOT use to pause and resume later (leave in-progress state alone if the user just wants a break), and do not use it to undo already-committed code changes — this skill stops orchestration state, it does not revert files.
- Rank 05View Agent SkillDirect
deep-dive
NotableTwo-stage front door for a fuzzy problem that might be a bug or might be a product question — run trace first to establish what is actually happening, then deep-interview to turn the confirmed problem into a spec. Use ONLY when the user explicitly asks for a deep dive or invokes /deep-dive, typically when they are not yet sure whether the issue is a defect or a scope question. Do NOT use when the nature of the problem is already clear: go straight to /trace for a known bug, or straight to /deep-
- Rank 06View Agent SkillDirect
deep-interview
NotableRun a Socratic interview that turns a vague request into a decision-complete spec, gated by a measured ambiguity score instead of a fixed question count. Use ONLY when the user explicitly asks for a deep interview, invokes /deep-interview, or asks to turn a fuzzy idea into a spec before planning. Do NOT use for ordinary implement/fix/debug requests, for questions with an obvious single answer, or as an automatic first step on complex-looking work — this skill never fires itself; the user or anot
Explore 1 more Agent Skills
- Rank 09View Agent SkillPossible
core-commit
NotableUse when committing work: stage by explicit path, write the project's commit format, check nothing private or unrelated is included. Never pushes unless you ask. Not for merging (use `core-phase-gate`) or opening a worktree (use `core-worktree`).
Frequently asked questions
- What kinds of skills are listed under Muse?
- Skills in this topic are SKILL.md packages that let an AI agent work with Muse, the AI-powered tool. They cover agent-side tasks such as connecting to Muse and driving its features from within an agent workflow.
- How is the list ordered?
- The list is ranked by observed search demand, with direct matches to the topic placed first. Any single publisher is capped so one source cannot fill the page.
- How do I pick one skill from the list?
- Match the skill's description to the task you need done, then read its SKILL.md for the dependencies and inputs it expects. Load one, run it against a real task, and add another only if you still have a gap.
- Do I need Muse set up before loading a skill?
- These skills operate on Muse, so the agent needs access to it for the skill to do anything. Check the SKILL.md for what the skill requires before you load it.
How to install an Agent Skill
Start with the first skill on this page. Claude Code reads a SKILL.md once it is in the skills directory:
- Claude Code — copy the SKILL.md for muse-delegation into ~/.claude/skills/muse-delegation/ for yourself, or .claude/skills/muse-delegation/ to share it with a project.
- From the registry — run npx skills add athola/claude-night-market@muse-delegation.
- Or download muse-delegation from its source and place that SKILL.md in the directory above.
How to choose Check the source and the skill’s own instructions before you install. Official and Notable describe where a skill comes from.