aktsmm/Agent-Skills

refine-product-100

プロダクトを反復実行で最高品質へ近づける。前回サマリーを読み、重複観点を避けて自律的に改善軸を選び、各回で具体改善・検証・記録まで行う。明示 release mode では品質 gate 後に配布と公開状態確認まで進める

Source text: Japanese

vscode-extension-guide

Guide for creating VS Code extensions and plugins from scratch through Marketplace publication. Use when developing a VS Code extension/plugin, adding commands or keybindings, building TreeView or Webview UI, publishing to Marketplace, or troubleshooting…

Source text: Mixed languages

document-knowledge-ingestion

Read multiple local documents deeply and preserve a source-linked Markdown knowledge library. Use for consolidating notes, books, PDFs, presentations, and screenshots across folders; recovering missed material; or turning a reference collection into…

sync-copilot-customizations

~/.copilot の instructions と agents を、公開 repo（Agent-Customization）の .github へ安全に同期する。syncToGlobal の分類と機密監査を通し、公開前に差分を提示して承認を取る。skill は対象外（private skills repo の retro / sync 系を使う）

Source text: Japanese

sync-public-skills

Synchronize curated Agent Skills across approved repositories with policy checks and verification. Use when publishing or updating managed skills, validating a repository mirror, or selecting a safe sync scope. Triggers on "sync public skills", "publish…

Source text: Mixed languages

retro-private-skills

Reflect reusable learnings into a private Agent Skills repository with scoped safety gates and same-run private commit/push. Not for editing ~/.copilot personal assets or workspace/repository assets (workspace and ~/.copilot skills may be read as intake…

retro-user

Reflect verified incident or session learnings into personal Copilot or Codex instructions and configuration. Invoke explicitly for Retro User, user-wide retrospective, 個人設定の知見反映, or global customization improvements. Not for workspace rules or private Skill…

retro-workspace

Reflect reusable learnings into the selected workspace's authorized instructions and automation assets, such as AGENTS.md, copilot-instructions.md, scoped instructions, scripts, and tasks. Use only for an explicit workspace/repo retro request: retro…

Source text: Mixed languages

review-workflow-structure

ワークフロー設計資産としての agent / instruction / skill / AGENTS を横断レビューし、SSOT・整合性・構造の問題を検出する

Source text: Japanese

session-handoff

Create a compact handoff note for another session. Use for セッションを閉じる, 引き継ぎを作る, 伝言を作る, create/export a handoff, or resume later. Do not create another handoff when a pasted note asks the current agent to resume work.

x-post-research

Research public X posts from keywords, hashtags, OR queries, domains, or known post URLs, then verify linked primary sources, official docs, repositories, benchmarks, and relevant media. Use for X で調べて, X の反応, ハッシュタグ調査, キーワードで X 検索, launch reactions, or…

x-twitter-browser-ops

X/Twitter browser analysis and explicitly approved mutations. Use for follower/following/mutual/verified analysis, X saved-post folder management, content/profile operations, or browser-driven troubleshooting. Not for public-post keyword/hashtag research or…

duck-critic

Run a Duck Critic producer-critic loop: you (main) keep producing the plan/code/tests and gate your own work at checkpoints with a different-model critic, revising until it passes. Use when asked for rubber duck, ラバーダック, 別モデルレビュー, second opinion, critic, or a…

biz-ops-setup

Set up or operate a business operations workspace for activity reports, tasks, and customer routing; workIQ is optional. Use when initializing BizOps, processing ongoing business updates, reviewing reports or tasks, or preparing an operations handoff.…

ai-cli-benchmark

Design, run, and review fair performance, token, and cost comparisons across GitHub Copilot CLI, Codex CLI, and direct model APIs. Use for AI CLI benchmark, Copilot vs Codex, prompt-cache diagnosis, cache read/write analysis, fresh-process latency…

Source text: Mixed languages

browser-bookmark-organizer

Safely audit, classify, deduplicate, and reorganize Chrome or Edge browser bookmarks/favorites while preserving toolbar shortcuts, active profiles, and observable sync state. Use for browser bookmark cleanup, duplicate bookmarks, folder redesign, or…

browser-max-automation

Visible browser automation using available Playwright, CLI, CDP, or computer-use tools with scoped actions and durable verification. Use for navigation, forms, screenshots, downloads, existing-session reuse, or iframe, modal, file chooser, passkey, and CDP…

Source text: Mixed languages

goal-loop

Run an explicit end-to-end goal loop with frozen Scope/criteria, worker delegation, external verification, evaluator review, and bounded retries. Use when the user asks for 'goal-loop', 'ゴールまで回す', '必ずやり遂げる', 'loop until done', 'until it actually works', or an…

permission-max

Reduce repeated permission prompts across Microsoft Scout, Copilot CLI, and host tool confirmations with user-approved settings and explicit before/after verification. Use when users ask for permission max, fewer approval prompts, shell allow-list expansion,…

Source text: Japanese

prompt-to-skill-reconcile

Prompt ファイル (.prompt.md) から Skill へ移行した後に、ワークスペース内の instructions / agents / README / Skill 本文 / スクリプトに残る旧 Prompt 前提の参照と運用手順を点検し、承認後に最小修正する。Use when: Prompt から Skill に移行した、migration 後の参照修正、.github/prompts 前提の記述が残っていないか確認、移行後の整合チェック。移行そのものの実行や Skill の新規設計には使わない。

Source text: Japanese

azure-infra-validation

Build and validate Azure infrastructure in a lab or sandbox using Azure CLI and official Microsoft docs. Use when creating verification environments, provisioning hub-and-spoke VNets, VPN Gateway or ExpressRoute-adjacent validation labs, checking…

Source text: Japanese

report-poc-to-requester

Use when the user asks to turn a finished or paused PoC / validation into a report for the requester: 依頼主向けレポート, 検証結果のご報告, 依頼元への返信, コピペできる報告文, summary-for-requester, 試したこと・結果・サマリ・証跡のフルパス. Builds a paste-ready report with conclusion first, results per item,…

Source text: Japanese

microsoft-forms-builder

Create and configure Microsoft Forms surveys through an owned browser session (Playwright over CDP): title, description, text/choice questions, free-text 'Other', required flags, external-respondent sharing, short URL and an anonymous-access check. Use for…

single-html-forge

Generate a single self-contained HTML file — a horizontal slide deck, a vertical-scroll explainer document, or a fixed-canvas summary image — with zero external runtime dependencies, then verify it mechanically. Use when the user asks for HTMLスライド,…

powerpoint-automation

Convert web articles or blog posts into slides through a content.json pipeline, translate or rebuild existing PPTX files onto a template, and edit an already-open PowerPoint deck with COM Automation (RefURL, overflow review). Use for article-to-slides…

Source text: Mixed languages

chrome-extension-dev

Chrome/ブラウザ拡張機能開発の包括的ガイド。WXTフレームワーク、Manifest V3、Chrome API、テスト手法、ストア公開（CWS API v2 での提出自動化を含む）をカバー。Use when: ブラウザ拡張機能を作成・修正・公開する時。Triggers on 'ブラウザ拡張機能', 'Chrome拡張', 'browser extension', 'WXT', 'content script', 'service worker', 'side panel', 'Chrome Web…

Source text: Japanese

skill-creator-plus

Create or review a reusable skill (SKILL.md) that packages a workflow, and decide whether the request should be a skill instead of a prompt, instruction, agent, or hook. Use when creating a new skill, extracting a workflow from a conversation, updating an…

agentic-workflow-guide

Design, review, and debug agent workflows, and decide when a request should use a prompt, instruction, skill, agent, or hook before escalating to multi-agent design. Use for .agent.md / .instructions.md / .prompt.md / AGENTS.md work, workflow architecture,…

azure-update-customer-pptx

Build a customer-facing Azure Update PowerPoint from Azure Updates MCP results, including customer classification, Japan region stamps, UPDATE Points, speaker notes, and Verify-Pptx gate checks. Use when creating or updating an Azure Update / Azure アップデート…

review-workflow-runtime

ワークフロー定義と参照 scripts/hooks/validators を横断し、実行時ロジック・I/O・冪等性・冗長処理・検証ゲートをレビューし、低リスク修正まで行う

Source text: Japanese

check-github-activity

現在または指定したGitHubリポジトリの新しいIssue、Pull Request、コメント、レビュー指摘をread-onlyで確認し、対応要否ごとに要約する。Use when: 新着Issue確認、新しいPR確認、コメント確認、レビュー返信確認、GitHub更新確認。

Source text: Japanese

code-review-multi-angle

差分、PR、選択範囲、指定ファイルを多角的にレビューし、既定は read-only、明示指示があれば critic review 後に修正と検証まで行う。日常コードレビューの入口で、`fix` 指定で Fix now / Guard now / Block に分けて最小修正・同根 Sweep・検証まで閉じる

Source text: Japanese

code-review-security

差分、PR、選択範囲、指定ファイルをセキュリティ観点で read-only レビューし、rubber duck / critic review で確認してから高信頼の脆弱性を報告する

Source text: Japanese

create-test-request

Use when: 会話・設計資料・回答文書から検証タスクを抽出し、別workspaceへ渡すテスト依頼文を作成する。受け手は最初に理解内容と実施項目を返し、承認後に検証する

Source text: Japanese

evaluate-waza-copilot-customizations

Waza で個人（~/.copilot）/ ワークスペース / 任意パスの instructions / agents / skills をセットアップ、static/runtime 評価、gate 単体実行、結果要約、report 更新まで行う。

Source text: Japanese

explain-release-changes

前回リリース以降の変更を、コードを知らない人にも伝わる説明文にする。Use when: リリースノートを一般向けに書き直したい、今回の変更を人に説明したい、ブログ/SNS/社内共有向けに変更点を紹介したい、機能の移植元（姉妹repo・参考OSS等）を明記して伝えたい

Source text: Japanese

export-copilot-session-dialogue

Interaction flow: セッション中の対話フロー・使用ツール・手順を構造化して出力。Use when: dialogue, tool flow, handoff, 手順再現。作業ログは export-session-log、再利用知見は export-knowledge を使う

Source text: Japanese

export-knowledge

Reusable learnings: セッション中の再利用知見を構造化し、ドメイン別（microsoft / copilot など）に振り分けて出力。Use when: knowledge, lessons learned, 知見化。作業ログは export-session-log、対話再現は export-copilot-session-dialogue を使う

Source text: Japanese

fact-check

ファイル、選択範囲、貼り付けた本文、URL 一覧に含まれる主張・数値・引用・参照先を 3 フェーズで検証する汎用ファクトチェック。

Source text: Japanese

generate-news-x-posts

Use when: 当日ニュース、ニュースURL、時事テーマから、事実確認済みのX投稿候補を複数作る。短い日常文や下書きのリライトは rewrite-x-post を使う

Source text: Japanese