メニュー
Load when writing or reviewing skills. Covers SKILL.md shape, loading mechanics, and how to structure bundled resources.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/haowjy/meridian-prompter --skill skill-artifactsコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Load when writing or reviewing agent .md files. Covers agent artifact shape, frontmatter, and Meridian-specific agent conventions.
Load when writing or reviewing prompts, skills, or agents. Gives the core heuristics for writing prompts that steer LLM behavior reliably.
Load when an agent, skill, or prompt needs strict review. Focuses on failure modes, severity, and clear feedback.
SOC 職業分類に基づく
SKILL.md を表示中
| name | skill-artifacts |
| type | reference |
| description | Load when writing or reviewing skills. Covers SKILL.md shape, loading mechanics, and how to structure bundled resources. |
| model-invocable | false |
Use /prompt-principles for broader prompt design and loading strategy. This skill covers how skill artifacts are structured in Meridian.
A skill is a directory under skills/ with a SKILL.md and optional resources/, scripts/, and assets/ subdirectories. The description is the trigger surface. The body is the loaded instruction layer. resources/ carry deeper content. scripts/ are for deterministic operations that should run instead of being re-derived in context.
Skills reach agents through four paths (see /prompt-principles/resources/skill-level.md for design guidance):
load: always in the system prompt. Use for skills the agent needs every run.available: agent can self-load when model-invocable: true. Use for skills it may reach for on demand.description: caller discovery surface. List attachable skill names so parent agents know what to compose via --skills.--skills: caller injection at spawn time. Bypasses available; any bundled skill can be injected.model-invocable defaults to true; an explicit model-invocable: false makes an available skill visible but not self-loadable (caller injection via --skills still works). Keep the global model-invocable pool small; each entry costs description tokens on every turn.
Set type: to declare the skill's role: principle for always-loaded thinking guidance, guardrail for always-loaded safety boundaries, mode-shift for on-demand shifts in activity, checkpoint for phase gates, reference for operational how-to.
Descriptions should lead with when to load the skill, not just what it contains. The description is what an agent or orchestrator reads to decide whether to load. "Load when reviewing test architecture" beats "Covers test structure patterns."