メニュー
Create tested project-local skills and scripts from demonstrated repeated work. Use after meditation or when a project workflow recurs and manual execution has caused inconsistency.
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。
SOC 職業分類に基づく
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/fabioc-aloha/Alex_ACT_Core --skill project-capability-authoring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。
SKILL.md を表示中
Preview, apply, or validate the curated read-in-place bridge from the global Copilot plugin store into Scout. Use after a Copilot plugin update, when Scout cannot resolve a bridged skill, or before adding a skill to the Scout allowlist.
Activates, verifies, repairs, or removes Core's 16 user-scope runtime instructions from canonical installed sources. Use after installing or updating Core, when Core identity or ACT behavior is inactive, or when Core receipt hashes drift.
Previews and applies Core's repository scaffold, portable workspace QoL settings, project Copilot settings, and Markdown CSS while preserving project-owned files. Use for new repositories, missing AGENTS/HANDOFF/episodic files, or project .vscode setup.
| name | project-capability-authoring |
| description | Create tested project-local skills and scripts from demonstrated repeated work. Use after meditation or when a project workflow recurs and manual execution has caused inconsistency. |
| lastReviewed | 2026-08-15T00:00:00.000Z |
Turn a demonstrated repeated project workflow into a small local capability. This skill creates project-local skills and scripts. It does not change Core, build MCP servers, or write persistent memory.
Do not use this for a one-off task, an untested aspiration, shared continuity, or a personal preference.
Prove the pattern: name the repeated workflow, the concrete failure or inconsistency it prevents, and the project users who benefit.
Scan prior art: search existing .github/skills/, scripts/, and
project guidance. Extend an existing capability when it already covers the
workflow.
Choose the smallest placement:
| Need | Destination |
|---|---|
| Guided repeatable workflow | .github/skills/<name>/SKILL.md |
| Workflow with deterministic automation | .github/skills/<name>/scripts/<task>.cjs |
| Cross-cutting project automation | scripts/<task>.cjs |
Define the contract: write the trigger, inputs, outputs, failure mode, and one should-fire plus one should-not-fire example. A script must state its prerequisites, accepted arguments, and validation command.
Preview the change: show every proposed file and the validation plan. Obtain explicit user approval before creating or modifying project files.
Validate: run the narrowest executable check for a script, and exercise the skill's should-fire and should-not-fire cases. Run diagnostics on every changed file.
Report: state the extracted pattern, artifact path, validation evidence, and why a new capability was warranted instead of a one-off solution.
Meditation may separately recommend user memory, repository memory, a handoff, or a chronicle. Do not write any persistent memory from this skill. A memory write requires the relevant explicit user request or approval and must pass the PII filter.
| Anti-pattern | Correction |
|---|---|
| Creating a skill after one routine task | Require a demonstrated repeat or concrete failure first. |
| Copying a skill that already exists locally | Extend the existing capability after the prior-art scan. |
| Hiding a destructive script behind a skill | Preview the command and require explicit approval before execution. |
| Creating project files while summarizing meditation | Meditation routes the candidate; this skill previews and validates it. |
| Treating an MCP server as a project script | Evaluate MCP construction separately; it is not this skill's scope. |
Revisit by 2026-11-15 if real meditations produce no demonstrated reusable workflow, authors repeatedly choose the wrong local placement, or generated capabilities fail their focused validation more than once.