MacHu-GWU

My personal, authoritative spec for building, developing, versioning, releasing, and wiring dependencies for my own Claude Code plugins through a split-repo plus central-marketplace setup. Use when creating a new plugin repo, laying out a plugin's .claude/skills/<name>/ folder, writing plugin.json or marketplace.json, tagging a plugin release, wiring plugin-to-plugin dependencies, authenticating private repos, or when I ask how to structure, release, or publish my Claude Code plugin.

Look up authoritative, up-to-date Google Antigravity documentation covering the Antigravity 2.0 IDE, the Antigravity CLI, the SDK, agents/subagents, models, skills, rules/workflows, hooks, MCP, browser use, settings, enterprise, migration, and FAQ. Use when the user asks how an Antigravity feature works, what a config field does, how to set up the IDE/CLI/SDK/skills/MCP, when troubleshooting an Antigravity error or unexpected behavior, or when you need to cite current official docs rather than rely on training-cutoff knowledge.

Look up authoritative, up-to-date OpenAI Codex documentation covering the CLI, IDE extension, Codex app, cloud, subagents, skills, MCP, sandboxing, configuration, enterprise, and integrations. Use when the user asks how a Codex feature works, what a config field does, how to set up the CLI/IDE/app/cloud/skills/MCP, when troubleshooting a Codex error or unexpected behavior, or when you need to cite current official docs rather than rely on training-cutoff knowledge.

Cross tool concept mapping knowledge base for AI coding agents. Use when you need to know how a project level configuration concept (project prompt, settings, skills, custom commands, hooks, MCP servers, subagents, permissions) corresponds across Claude Code, Codex, and Antigravity, or how to port such a setup from one tool to another.

Read-only review of how completely a project's Antigravity configuration has been ported to Claude Code. Scans both sides, compares them against the coding-agent-concept-mapping knowledge base, and writes a gap report to tmp/review-port-antigravity-to-claude-code.md. Makes no changes to the project config. Use to audit a Antigravity to Claude Code migration.

Port a project's Antigravity configuration to Claude Code, keeping the Antigravity files in place. Scans the project for Antigravity config artifacts, maps each to its Claude Code equivalent using the coding-agent-concept-mapping knowledge base, and creates or edits the matching Claude Code files. Use when migrating a repo from Antigravity to Claude Code.

Read-only review of how completely a project's Antigravity configuration has been ported to Codex. Scans both sides, compares them against the coding-agent-concept-mapping knowledge base, and writes a gap report to tmp/review-port-antigravity-to-codex.md. Makes no changes to the project config. Use to audit a Antigravity to Codex migration.

Port a project's Antigravity configuration to Codex, keeping the Antigravity files in place. Scans the project for Antigravity config artifacts, maps each to its Codex equivalent using the coding-agent-concept-mapping knowledge base, and creates or edits the matching Codex files. Use when migrating a repo from Antigravity to Codex.

Look up authoritative, up-to-date Claude Code, Agent SDK, hooks, MCP, skills, plugins, settings, subagents, and admin documentation. Use when the user asks how a Claude Code feature works, what a config field does, how to set up hooks/MCP/skills/subagents/plugins, when troubleshooting a Claude Code error or unexpected behavior, or when you need to cite current official docs rather than rely on training-cutoff knowledge.

Maintain the release-history.rst changelog for a Python project before publishing a new version. Use when the user wants to prepare release notes, update changelog, or is about to publish/tag a new version.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

File-based task management with shai-tix CLI. Use when user mentions kanban, ticket, planning, task tracking, or references tickets by [id] or #id format. MUST run via: uvx --from shai-tix==0.1.3 shai-tix <cmd>. Data stored in .tix/ directory (NOT .shai-tix/).

Create properly formatted references to paths (files or folders) in a git repository. Use when writing docs that need to link to any path in the repo.

Python unit testing patterns for pytest, including test file structure, fixtures, mocks, and coverage strategy. Use when asked to write tests (unit/integration) for any Python module, function, class, or method.

Use when exploring Python project structure, finding virtualenv paths (python/pip/pytest), or locating package metadata (name, version, pyproject.toml).

Find and add authoritative source links for all facts, citations, and references in markdown documents

Foundation skill for authoring teaching repositories (hands-on IT learning courses). Defines the shared concept model, language/format conventions, and document templates that the whole lesson-smith-* skill family builds on. Load when creating or updating any course material — README.md, README-cn.md, TICKET.md, CLAUDE.md, TEACHING.md, Syllabus.md, about.md.

给一个 upskill 教学仓库收尾定型. 重写 README-ORIGINAL (Lesson 门面), 写出根目录的 README 与 TICKET (仓库总览加整门课验收清单, 一次产出全部语种), 再跑 lesson-smith sync 生成 SYLLABUS 与快照, 最后跑 lesson-smith lint 把关. 创作流的最后一步, examples 与 forge 都做完之后用. 关键词, finalize upskill, 重写 README-ORIGINAL, 写根 README TICKET, 生成 SYLLABUS, upskill 收尾.

On-call learning mentor for this upskill course. Walks you through the examples progression, unpacks any specific file or spot you are stuck on, and helps you decide what to learn next. Auto-loads when you mention or open any file under examples/, or say things like "walk me through this course", "explain this example", "I'm stuck here", "what should I learn next". Also invocable directly.

Interview-style quiz that checks whether you have truly internalized this upskill course (know-what and know-why). Auto-loads when you say "quiz me", "test my knowledge", "let's drill", when you mention the quiz, or when you mention the quiz mini task file under examples/. Also invocable directly.

带教程创作者从零创作一个符合 upskill 规范的 repo, 主要是 examples 下那一整套课程内容. 讨论规划, 试水前几篇锁风格, 迭代往后写, 收尾写 quiz 与梳理, 批量翻译, 最后交给 forge. 创作或续写 upskill 课程时用. 关键词, 写 upskill 课程, 创作 examples, 规划这门课.

为当前这个 upskill 教学仓库锻造出 "学这门课" 的工具链. 扫一遍 examples 渐进序列, 向创作者问清学习素材与考法, 写出 docs/upskill/ 下的 3 份文档 (learn, runbook, quiz), 并落下 upskill-learn 与 upskill-quiz 两个子 skill. 接手或刷新一个 upskill repo 的学习工具时用. 关键词, forge upskill, 生成 upskill 学习 skill, 刷新 upskill 文档.

How Agent Skills and Subagents differ and how to combine them — the skill-as-caller / subagent-as-doer best practice. Use when deciding whether a task should be a skill or a subagent, or when authoring a skill and subagent that work together (e.g. a /name skill that delegates to an isolated subagent).

Guide for building Alfred workflow Python packages using the afwf SDK. Use when creating, modifying, or debugging afwf-based Alfred workflows — Script Filters, CLI entry points, fuzzy matching, disk caching, write actions, testing, or deployment.

Maintain the release-history.rst changelog for a Python project before publishing a new version. Use when the user wants to prepare release notes, update changelog, or is about to publish/tag a new version.

Teaches how to use the sayt2 Python library to build search-as-you-type applications. Use when the user asks how to use sayt2, wants to build a search index, or needs help with sayt2 DataSet, fields, queries, sorting, or caching.

Maintainer guide for atlas_doc_parser development. Use when implementing ADF marks/nodes, understanding codebase architecture, or writing tests.

Fetch raw ADF JSON data from a Confluence page URL. Use this skill when you need to see real-world ADF examples, understand how Confluence represents specific elements, debug ADF parsing, or create test samples.

Query Atlassian Document Format (ADF) JSON schema definitions to understand ADF node and mark types. Use this skill when implementing ADF dataclass nodes/marks, or when user asks about ADF structure, ADF nodes, ADF marks, or Atlassian Document Format implementation.

Punctuation convention for mixed Chinese and English documents. Use when writing or revising any .md file whose narrative is mainly Chinese but whose technical terms stay in English, to keep every punctuation mark in its English (ASCII) form rather than the Chinese full width form. Trigger phrases include "写中文文档", "中英混排", "标点规范", "normalize punctuation", "fix Chinese punctuation", or any request that produces or edits Chinese narrative with English technical terms.

Markdown writing conventions for project documents. Use when authoring or revising any .md file in this project (meeting notes, project docs, design notes, READMEs, tickets) to keep heading structure, punctuation, and tone consistent. Trigger phrases include "write a doc", "draft a markdown", "format this note", "tidy up this markdown", or any request that produces or edits a .md file.

Translate/rewrite files into English. Supports PDF, Markdown, plain text, and other formats. Generates an -en.md file in the same directory. Use when user wants to translate documents into English for easier reading.

How to use the boto_session_manager library. Use when writing Python code that interacts with AWS via boto3, creating boto sessions, assuming IAM roles, managing AWS credentials, or using typed AWS service clients. Trigger when: code imports boto_session_manager, user mentions BotoSesManager, or user asks how to use bsm.

Interactive guide to help maintainers learn the boto_session_manager project. Use when someone asks how the project works, wants to understand the architecture, or is onboarding as a new maintainer.

Review a draft Chinese product slogan saved on disk, give 2-3 sharp and actionable improvement suggestions, and save them as a markdown file under tmp/. Use when the user asks to critique or 评审 a slogan file.

Draft a short Chinese product slogan (no more than 15 Chinese characters) and save it as a markdown file under tmp/. Use when the user asks to write or revise a slogan / tagline for a (usually fictional) product.

Guide for building Alfred workflow Python packages using the afwf SDK. Use when creating, modifying, or debugging afwf-based Alfred workflows — Script Filters, CLI entry points, fuzzy matching, disk caching, write actions, testing, or deployment.

Generate one secure password, short ID, or UUID4 from the shell via the afwf-genpass CLI (the `*-one` subcommands). Use when the user asks for a random password, random ID/slug/token, or UUID4 on the command line — or wants to pipe one into another shell command, env var, or script.

Guide for using the func_args Python library (>=1.0.2) — sentinel-based function argument handling with REQ/OPT markers and enhanced dataclasses. Use when writing API wrappers, parameter validation, or dataclass-based config objects.

Translate/rewrite files into English. Supports PDF, Markdown, plain text, and other formats. Generates an -en.md file in the same directory. Use when user wants to translate documents into English for easier reading.