shunk031
User💻 My dotfiles powered by chezmoi; built around Rust-based tools (sheldon/starship/mise, etc.) featuring a unified coding-agent setup across Codex, Claude Code, and Gemini.
Categories
Indexed Skills (25)
cgd-dev-identity
Use this skill for all GitHub write operations in creative-graphic-design organization repositories (such as design-generators), including git commits and pushes, PR creation and updates, issue and PR comments, labels, milestones, and user-requested merges. Perform them as the creative-graphic-design-dev machine user. Use it both when the user explicitly requests the bot or dev account and whenever beginning write work in an organization repository without an explicit request.
transformers-convert
Use this skill when converting custom PyTorch models to Hugging Face Transformers format. Helps with: (1) Creating PretrainedConfig and PreTrainedModel classes, (2) Writing ImageProcessor/Tokenizer, (3) Compatibility testing, (4) Hub upload preparation. Use when the user wants to make their model compatible with transformers library.
gh-comment-attach-files
Attach local files to a GitHub issue or pull request comment via Playwright CLI and return the hosted attachment URLs without submitting the comment. Use when you need GitHub-hosted image or document URLs on github.com or GitHub Enterprise Server.
high-impact-journal-publishing
Guidance and workflow for planning, writing, submitting, revising, and publishing scientific manuscripts in high-impact journals. Use when Claude needs to advise on research quality, study design, manuscript structure, journal selection, peer review responses, or publication strategy.
manage-agent-guidance
Organize persistent coding-agent guidance and its adapters without blindly editing a user-named file. Use when creating, converting, reviewing, or reorganizing AGENTS.md, CLAUDE.md, user-level instructions, custom-agent wrappers, or repo-local skills; when deciding whether a rule belongs at user, repository, subtree, task, or skill scope; or when preventing duplicated guidance across tools. Always establish scope, source of truth, and existing ownership before drafting or applying a persistent rule.
structured-writing
Structure and revise detailed instructions, plans, reports, documentation, and bullet lists so their hierarchy and relationships are clear. Use when drafting or reorganizing multi-part prose, especially when deciding between headings, topic bullets, supporting nested bullets, and independent flat items. Do not use for short answers, code-only output, or machine-readable data unless the user also requests prose organization.
shunk031-cgd-dev-identity
Use this skill for all GitHub write operations in creative-graphic-design organization repositories (such as design-generators), including git commits and pushes, PR creation and updates, issue and PR comments, labels, milestones, and user-requested merges. Perform them as the creative-graphic-design-dev machine user. Use it both when the user explicitly requests the bot or dev account and whenever beginning write work in a creative-graphic-design organization repository without an explicit request; never use it in repositories of any other organization or host.
shunk031-gh-comment-attach-files
Attach local files to a GitHub issue or pull request comment via Playwright CLI and return the hosted attachment URLs without submitting the comment. Use when you need GitHub-hosted image or document URLs on github.com or GitHub Enterprise Server.
shunk031-high-impact-journal-publishing
Guidance and workflow for planning, writing, submitting, revising, and publishing scientific manuscripts in high-impact journals. Use when Claude needs to advise on research quality, study design, manuscript structure, journal selection, peer review responses, or publication strategy.
shunk031-manage-agent-guidance
Organize persistent coding-agent guidance and its adapters without blindly editing a user-named file. Use when creating, converting, reviewing, or reorganizing AGENTS.md, CLAUDE.md, user-level instructions, custom-agent wrappers, or repo-local skills; when deciding whether a rule belongs at user, repository, subtree, task, or skill scope; or when preventing duplicated guidance across tools. Always establish scope, source of truth, and existing ownership before drafting or applying a persistent rule.
shunk031-structured-writing
Structure and revise detailed instructions, plans, reports, documentation, evidence documents such as TRAINING.md, repository protocol documents, pull-request bodies, and bullet lists so their hierarchy and relationships are clear. Use when drafting or reorganizing multi-part prose and repository completion or status reports, even when the resulting report should be concise, especially when deciding between headings, topic bullets, supporting nested bullets, and independent flat items. Do not use for ordinary conversational short answers, code-only output, or machine-readable data unless the user also requests prose organization.
shunk031-transformers-convert
Use this skill when converting custom PyTorch models to Hugging Face Transformers format. Helps with: (1) Creating PretrainedConfig and PreTrainedModel classes, (2) Writing ImageProcessor/Tokenizer, (3) Compatibility testing, (4) Hub upload preparation. Use when the user wants to make their model compatible with transformers library.
shunk031-herdr-tab-status
Choose and update the current Herdr worker tab name with exact leading status emojis for active work, blocked work, user handoffs, and completion. Use whenever a Herdr tab label or its status meaning must be chosen or updated.
shunk031-orchestrate-herdr-workers
Orchestrate and coordinate coding workers in Herdr by spawning parallel Codex workers in git worktree tabs, tracking status and reports, and routing owned task, review, and pull-request lifecycles. Use only when the request explicitly mentions Herdr and asks to orchestrate, coordinate, fan out, parallelize, delegate, or route work among coding workers — or a prompt tells you that you are a Herdr worker reporting to an orchestrator. Never use it merely because a task involves parallel work, multiple agents, or git worktrees without Herdr being named; also not for single-agent Herdr control or general Herdr CLI questions, which the herdr skill covers.
shunk031-doc-slop-review
Review reader-facing text with the repository's two-tier slop review before publishing it. Use when writing or editing documentation, README or TRAINING files, issue bodies and comments, pull request bodies and descriptions, or status reports that another person will read. Runs deterministic checks and one blind model judge over the draft, quotes each problem, and produces a PASS or FAIL to attach to the publish report; an unavailable review is a failed gate, not a PASS.
ai-slop-checklist-ja
AI臭い日本語をレビューし、書き手の不在、主体の弱さ、構造テンプレ、抽象語、翻訳調、表層記号をチェックして修正案まで出すスキル。ブログ記事、技術記事、note/Zenn草稿、SNS投稿、メール、README、発表原稿などの日本語を、AIっぽさの診断、採点、具体的な直し方、軽いリライト案まで含めて確認したいときに使う。
humanizer-ja
AIくさい日本語を、人が書いたように自然でこなれた日本語へ書き換えるスキル。ブログ記事、メモ、メール、文書、SNS投稿、チャット返信などの日本語を書いたり推敲したりするときに使う。AIっぽい言い回しを消したいとき、日本語を自然にしたいとき、言い換えたいとき、硬さを抜きたいとき、トーンを整えたいとき、人間らしい声や温度を足したいときに使う。
python-uv-workflow
Apply Python development policy using uv-first execution, test-first behavior validation, and pre-commit quality gates. Use when implementing or refactoring Python code.
setup-agent-docs
リポジトリを coding agent 対応にするときの agent ドキュメント配線パターン(AGENTS.md を source of truth にし、CLAUDE.md は symlink にする)。新しいリポジトリのエージェント対応、coding agent ready 化、AGENTS.md / CLAUDE.md の新規作成・変換・整理、repo-local skill の設置、「エージェント向けドキュメントを整備して」という依頼のときに必ず使う。CLAUDE.md だけ・AGENTS.md だけを作ろうとしている場合もこの skill で配線を確認する。
shdoc-shell-docs
Write and review shellscript documentation with shdoc annotations. Use when a coding agent creates, edits, or reviews `.sh` files or shell executables and should add, repair, or normalize `@file`, `@brief`, `@description`, `@arg`, `@option`, and `@example` comments to match shdoc conventions.
shunk031-ai-slop-checklist-ja
AI臭い日本語をレビューし、書き手の不在、主体の弱さ、構造テンプレ、抽象語、翻訳調、表層記号をチェックして修正案まで出すスキル。ブログ記事、技術記事、note/Zenn草稿、SNS投稿、メール、README、発表原稿などの日本語を、AIっぽさの診断、採点、具体的な直し方、軽いリライト案まで含めて確認した��ときに使う。
shunk031-humanizer-ja
AIくさい日本語を、人が書いたように自然でこなれた日本語へ書き換えるスキル。ブログ記事、メモ、メール、文書、SNS投稿、チャット返信などの日本語を書いたり推敲したりするときに使う。AIっぽい言い回しを消したいとき、日本語を自然にしたいとき、言い換えたいとき、硬さを抜きたいとき、トーンを整えたいとき、人間らしい声や温度を足したいときに使う。
shunk031-python-uv-workflow
Apply Python development policy using uv-first execution, test-first behavior validation, and pre-commit quality gates. Use when implementing or refactoring Python code.
shunk031-research-before-implementation
Research current official web documentation and representative GitHub implementation code before designing or editing non-trivial work involving third-party tools, libraries, platforms, APIs, configuration formats, or version-dependent behavior. Use for implementation, migration, integration, and configuration tasks where local files or memory alone cannot establish current supported behavior.
shunk031-shdoc-shell-docs
Write and review shellscript documentation with shdoc annotations. Use when a coding agent creates, edits, or reviews `.sh` files or shell executables and should add, repair, or normalize `@file`, `@brief`, `@description`, `@arg`, `@option`, and `@example` comments to match shdoc conventions.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.