dots

Step-by-step article writing workflow with style selection (personal voice vs serious), a hallucination-prevention fact-check gate, and AI-smell proofreading. Use when writing or co-writing blog/tech articles — triggers: 「記事を書く」「ブログ記事」「記事化して」, turning memos/logs/repos into an article, or drafting for funailog/Zenn/note. For proofreading-only requests, use the japanese-ai-writing-proofreader skill directly.

Proofread Japanese prose in three passes: textlint (mechanical), AI-smell removal, and deep naturalness (sentence-end variety, rhythm, connectives). Use when asked to 「校正して」「推敲して」「AI臭を消して」「自然な日本語にして」, when polishing article drafts / PR bodies / docs written in Japanese, or as the proofreading phase of the article-writing skill.

Iteratively improve agent-facing text instructions (skills, slash commands, task prompts, CLAUDE.md sections, code-generation prompts) by having a bias-free executor run them and evaluating from both sides (executor self-report + caller-side metrics). Repeat until improvement plateaus. Use immediately after creating or significantly revising a skill/prompt, or when unexpected agent behavior is suspected to stem from ambiguous instructions.

Design where a repository's agent knowledge lives — which conventions belong in which directory-level CLAUDE.md, and which "convention-definition" skills should be copied into CLAUDE.md and then deleted. Optimizes for what the model gets wrong without the note, not documentation coverage. Loads the session model's profile from ~/.claude/references/model-profiles/ to match its writing style. Use when onboarding a repo to Claude Code, after a model upgrade, or when CLAUDE.md has grown bloated or conventions are scattered across skills. Triggers: 「CLAUDE.md を設計」「階層別 CLAUDE.md」「作法を CLAUDE.md に移したい」.

Use when the user wants to set up a recurring or autonomous agent loop and needs to design it before running it. Triggers: 「ループを設計」「自律ループを 作りたい」「loop を組みたい」, turning a vague "I want a loop that does X" into a concrete spec, or choosing a runner (/loop, /schedule, ralph-loop, pr-monitor) for repeated work. NOT for running an already-designed loop, and NOT for one-off tasks.

Ship a GitHub Issue end-to-end: fetch the issue as the spec, implement via a worktree-isolated subagent, open a PR that closes the issue, then watch CI through merge. GitHub Issues is the single source of truth.

Watch a GitHub PR asynchronously until CI is green and review threads are resolved (or through merge), using the harness Monitor tool so the main session keeps working. Optional auto-fix mode reacts to CI failures by fixing and pushing on the PR branch. Use when (1) the user asks to watch a PR or CI (「PR見てて」「CI監視して」), (2) changes were just pushed and CI needs confirmation, (3) another skill (e.g. issue-ship) needs CI watched through merge. Self-contained — no plugin dependency, so it works on every machine.

E2E testing workflow for web applications using Chrome in Claude. Automates functional testing, bug detection, and GitHub Issue creation. Use when: (1) running functional tests on web apps, (2) verifying staging/production behavior, (3) finding bugs and creating Issues, (4) capturing UI behavior with screenshots

End-to-end workflow for finding content, app, service, or product opportunities where search demand is high but quality supply is low (trend arbitrage). Use when: finding content gaps, keyword opportunities, niche research, analyzing whether a topic or idea is worth pursuing, or competitive gap analysis.

Create viral X/Twitter posts based on the official X algorithm (Phoenix/Grok). Generates and improves posts optimized for maximum engagement. Use when (1) creating X posts for products, services, or information, (2) improving existing post drafts, (3) designing thread structures for long content.

Zero-cost scaling patterns for personal/indie web apps that may go viral: client-side processing (ONNX Runtime Web via WebAssembly/WebGPU), no user-data storage, free static hosting (Cloudflare Workers), asset optimization, UI simplicity as a feature. Use when (1) designing or reviewing the architecture of a personal/indie web app, (2) choosing hosting or infrastructure for a side project, (3) the user asks how to keep infra cost at zero or survive viral traffic spikes.

Guided feature development with harness engineering principles. Sprint contracts, architecture design, implementation, evaluation, and quality review. Use when building new features or significant enhancements.

Conduct an attacker-perspective security review of a codebase, then ship the fixes as multiple non-conflicting pull requests in parallel using subagents and git worktrees. Verifies findings to filter false positives, plans fixes via file-occupation analysis to guarantee no merge conflicts, dispatches each fix to its own TDD-driven subagent in an isolated worktree, then pushes and opens one PR per fix. Use this whenever the user asks for a security audit, pentest, vulnerability review, attack surface analysis, or wants to "find vulnerabilities" / "review for security issues" in their own (or otherwise authorized) project, even if they don't say the word "pentest". Also use when the user already has a list of independent fixes (security or otherwise) and wants them shipped as parallel non-conflicting PRs without manually managing worktrees.

Generate a 1-page purchase memo for a US or JP individual stock (ETF support in v1.1). Walks 5 phases — business, fundamentals, valuation, risks, verdict — and ends with deep-dive options. Use when (1) the user asks "should I buy <TICKER>?", (2) the user wants a fundamental review of a specific company, (3) the user pastes a 10-K / 有報 / IR URL and asks for an analysis. Skip for ETF / index / dividend-focused questions in v1.

Generate an IPO subscription-decision memo (抽選参加判断) for a Japanese pre-listing IPO. Walks 5 phases — business, fundamentals, supply/demand, risks, verdict — where supply/demand is the protagonist. Use when (1) the user asks "このIPO参加すべき？", (2) names an upcoming JP IPO / its ticker, (3) pastes a 目論見書 / 有価証券届出書 URL. v1 = JP pre-listing only (US S-1 / just-listed / pre-IPO unlisted are out of scope).

Research a repo/docs in parallel for a given topic and generate a disposable GFM+mermaid learning primer (Markdown). Triggers (Japanese): 「〜の学習資料を作って」 「オンボーディング資料がほしい」「〜をキャッチアップしたい」「primer 作って」. Output is not committed (defaults to repo-local `.ai/`, globally git-ignored); view with mo or any GFM+mermaid viewer.

Delegate an interrupt task to a background agent in an isolated worktree. The main session continues uninterrupted. The agent implements, commits, pushes, and creates a PR. Use when a quick fix or small task needs to happen without losing current context.

Pre-submission review for iOS App Store. Scans the codebase for common rejection reasons and generates a pass/fail report with fixes. Use this skill when the user mentions App Store submission, review, release, 審査, 提出, or phrases like "ready to submit", "before submitting to Apple", "submission review", "rejection check". Also trigger when the user is preparing a TestFlight build for external review or discussing App Store rejection issues. Supports Swift/SwiftUI, UIKit, and React Native projects with Apple IAP and RevenueCat.

UI design quality standards and principles for frontend implementation and code review. Use when (1) implementing UI from design specs or mockups, (2) reviewing frontend/UI code, (3) creating new UI components, (4) building user interfaces for web or mobile apps. Complements frontend-design skills with quality enforcement.