一键导入
skill-name
Brief description of when to use this skill
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Brief description of when to use this skill
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Issue-driven development with worktree isolation. 5-stage pipeline (new, plan, work, review, merge) with human gates between stages. Use when creating tasks, planning, implementing, reviewing, or merging.
Start or join an inter-agent chat session. Two Claude Code sessions communicate via a shared markdown file in tmp/. Use when the user says "ai-chat", "start a chat", "agent chat", or provides a chat slug to join.
View and manage configurations and skills for AI agent CLI tools (Claude Code, Opencode). Use when checking config files, listing or managing skills, viewing model settings, comparing configs across tools, or performing common config operations.
Orchestrate external AI agent instances (Claude Code, Opencode, Gemini CLI) via SDK or CLI. This skill NEVER does work itself — it always delegates to an external agent.
Parse Word documents (.docx) into repository-friendly markdown artifacts. Use when users need to extract content from Word files, convert to markdown, or batch process document directories.
Parse Power BI files (.pbix) into repository-friendly JSON and CSV artifacts. Use when users need to extract metadata, measures, or data model structure from Power BI files.
| name | skill-name |
| description | Brief description of when to use this skill |
[What this skill does and when to use it. Commands and agents should reference this skill directly.]
[Prerequisites, user input expected, and how context will be provided. Keep this delivery-agnostic: describe what inputs the skill needs (files, parameters), not where they come from. The app layer (CLI, web UI, API) handles how files arrive — the skill should work with whatever file paths are provided.]
[Optional - only for parent skills that have sub-skills]
CRITICAL: You MUST load the appropriate sub-skill from the sub-skills/ directory when routing is needed.
[Optional - if skill has supporting files. These are read on demand, not loaded with the skill.]
[Principles to follow when executing this skill]
[Optional - for domain-specific information that doesn't fit above]
[Use this section for additional detailed information such as architecture patterns, naming conventions, integration patterns, or reference commands.]