メニュー
Manifest a new hierarchy of modular skills following the Late-Binding Hierarchical Skill Framework.
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/susamn/aistuff --skill skill-manifestor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。
Manage the dotfiles repository, stow packages, and configure agent skills. Use when adding a skill, modifying the stow setup, or working inside ~/dotfiles.
Practice LeetCode problems in a 3-tab mosaic dashboard (Problem, Intuition, Solution — Python + Go). Starter-seeded with 100 classic problems; add, swap, or remove any problem by URL or description. Use when asked to practice LeetCode, add/swap a coding-interview problem, or open the LeetCode practice dashboard.
Import external skills from Git repositories, local directories, or web URLs into the local skills library, adapting them to house conventions, auditing, and suggesting cross-skill integrations. Use when importing a skill, bringing in an external skill, or adding a skill from GitHub.
SOC 職業分類に基づく
SKILL.md を表示中
| name | skill-manifestor |
| description | Manifest a new hierarchy of modular skills following the Late-Binding Hierarchical Skill Framework. |
| version | 2.0.0 |
| kind | guidance |
| triggers | ["manifest a new skill hierarchy","create skills for [problem]","use skill-manifestor"] |
| intent | meta-skill |
| type | action |
| parent | null |
| mode_aware | true |
| created_at | 2026-05-30T00:00:00.000Z |
| updated_at | 2026-07-29T00:00:00.000Z |
This skill is the architect for the Late-Binding Hierarchical Skill Framework. It is used to design and manifest new skill hierarchies from a root parent level down to implementation-specific children.
Trigger: "Manifest a new skill hierarchy" or "Create skills for [Problem]."
When triggered, the Agent acts as an Architectural Blueprint to construct a new ecosystem of modular skills.
The Agent MUST ask the following questions before generating any files:
[namespace]-testing, [namespace]-store, etc.)[skill-id] inside the flat skills/ root.SKILL.md within that directory (e.g., skills/[skill-id]/SKILL.md).[namespace]-parent.[namespace]-[category] or [namespace]-[category]-[sub-category] for deeper specialization.Every generated skill MUST include the following YAML frontmatter in its SKILL.md:
name: [skill-id]
id: [skill-id]
version: 1.0.0
type: [context | action | validation]
[ ]
Each SKILL.md MUST contain two distinct sections: ## [TOOL_MODE] and ## [LIBRARY_MODE].
Once the hierarchy design is finalized, the Agent MUST use skill:skill-creator to handle:
mkdir -p skills/[skill-id]).SKILL.md file generation with the finalized content../do-stow.sh to make the skills available across all agents.[namespace]-parent)This SKILL.md MUST contain the Global Laws for the domain, including:
CONTEXT.md: Explicitly defines the module's primary responsibility and scope.TODO.md: A live task list. Items must be ticked off only upon completion and verification.ARCHITECTURE.md: Contains the project architecture, requirements, and high-level design.DEPENDENCY.md: Clearly defines inter-module dependencies. Rule: This file must be symlinked into every child module.PROJECT_STATE.md: Mandatory root file to track architectural decisions and progress. Rule: This file must be symlinked into every child module.
- **Phase**: [Current Phase]
- **Type**: [Fullstack/Frontend]
- **Stack**: [List of selected techs]
- **Modules**: [List of modules]
- **API Status**: [Draft/Finalized]
## Module Updates
[Each module adds its high-level status and major changes here, only relevant and single line containing the above information so that it can be easily read by other modules for their Reference]
TODO.md at a time.These SKILL.md files focus on specific layers (e.g., webapp-testing) or sub-specializations (e.g., webapp-testing-bdd):
parent: metadata and skill:[parent-id] in text for global laws.webapp-testing-bdd) MUST refer to its category parent (webapp-testing), which in turn refers to the root parent (webapp-parent).[namespace]-vulnerability)If applicable to the domain, this skill MUST be included to handle security integrity:
webapp, IAM misconfigurations for cloud, buffer overflows for system).Trigger: "Audit the [namespace] skills" or "Validate [Project] against skill framework."
When triggered for skills, the Agent acts as a Recursive Auditor for the skill library itself:
skills/[skill-id]/SKILL.md pattern.SKILL.md has a valid YAML header with correct parent pointers.[TOOL_MODE] and [LIBRARY_MODE] blocks.skill:<name> references correctly point to existing SKILL.md files.When triggered for a codebase, the Agent enforces the Global Laws defined in the hierarchy:
CONTEXT.md or TODO.md.ARCHITECTURE.md and DEPENDENCY.md (and check for symlinking in modules).OpenAPI schema exists and is being utilized as the source of truth.skill:webapp-vulnerability), STOP and request it.skill:<name>)To resolve a logic gap, the Agent MUST:
skill: prefix.read_file tool to fetch skills/[name]/SKILL.md.[MODE] block into the current reasoning context.