Get a branch ready for the maintainer's final review: review it for security, simplicity, and billing abuse, fix what's real, get checks passing, and open or update the PR. Use whenever the user says a feature, fix, or branch should be "merge ready", asks to get changes ready for review, or appends this to a build request ("build X and make it merge-ready").
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Get this branch ready for the maintainer's final review. If you were also asked to build something, build it first.
Fetch and merge origin/main if the branch is behind. Then review everything on the branch against it, including uncommitted and untracked files. Make sure you understand every product decision and change in functionality.
Is it secure?
Is it as simple as it can be? What can be deleted or simplified given the end state of the branch? Did an unlikely edge case lead to overengineering? We prefer to delete code.
Does it use APIs and SDKs (TanStack, DataForSEO, Autumn, Cloudflare) idiomatically? Watch for code that reinvents what they already do, works around limits they don't have, or pays for more than it needs. Check the docs only when something looks off.
Is there a significant way to abuse billing, like getting credits, paid features, or provider calls without paying? Ignore small leaks nobody would bother exploiting.
Beyond that, review it however you think best. Work out which parts could do the most damage if they're wrong and spend your time there. docs/maintainers/review-guidelines.md lists the repo's invariants and the things that look wrong but are intentional.
Fix what's real. Commit as you go. Get pnpm ci:check and pnpm test passing locally, then push. Those local runs stand in for CI, so don't wait for GitHub CI to finish.
Open or update the PR against main with a high-level description: what changed and why, what to look at first, and anything you left for the maintainer's judgment. No per-file changelog.
Never merge. The maintainer reviews last.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Find the questions people ask about a market, how ChatGPT answers them, and which sites get cited in the answers. Use when the user asks what people ask AI about their niche, wants prompt ideas, or wants to know where their brand is missing before choosing prompts to track. Research only; it never saves tracking or starts answer collection.
日本語の概要は準備中です。原文の説明を表示しています。
Audit how a brand shows up in AI answers about its market and deliver a short report on the few changes most likely to get it mentioned or cited, such as a third-party page to get onto, an owned page to improve, or an access problem to fix. Use when the user asks for an AI visibility audit, a GEO or AEO audit, why AI recommends competitors instead of them, or how to show up in ChatGPT, Gemini or Google AI answers.
日本語の概要は準備中です。原文の説明を表示しています。
Map SEO market leaders, winning content themes, keyword coverage, backlinks, and strategic gaps.
日本語の概要は準備中です。原文の説明を表示しています。
Analyze one competitor's organic footprint, ranking keywords, content themes, backlinks, and gaps.
日本語の概要は準備中です。原文の説明を表示しています。
Create or update a skill in this repository the right way — canonical home in .agents/skills, internal-vs-public marking, symlink mirroring into .claude/skills, and public docs registration for product skills. Use whenever adding a new skill, converting a workflow into a skill, or when .claude/skills and .agents/skills look out of sync.
日本語の概要は準備中です。原文の説明を表示しています。
Remove AI writing patterns from prose so it reads like a person wrote it. Use when writing, editing, or reviewing user-facing text (blog posts, docs, marketing copy, PR descriptions), or when asked to "deslop", "de-AI", or "make it sound human".
日本語の概要は準備中です。原文の説明を表示しています。