菜单
Use when Orlix work benefits from explicit workflow packets, authorized subagent orchestration, or independent research, implementation, test, and review tracks.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
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。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/rudironsoni/Orlix --skill orlix-dynamic-workflows命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | orlix-dynamic-workflows |
| description | Use when Orlix work benefits from explicit workflow packets, authorized subagent orchestration, or independent research, implementation, test, and review tracks. |
Use orchestration when work has independent tracks, crosses ownership boundaries, needs separate skeptical verification, or the user explicitly requests delegation. Do not spawn agents unless the user or current environment authorizes it.
Store reusable workflow mechanics in the owning skill and current packet state in structured task envelopes under Build/AgentHarness/. Do not create WORKFLOW.md, plan journals, or implementation logs.
Every packet names its objective, required canonical pages and reports, owning layer, allowed and forbidden scope, expected output, verification, and integration point. Keep packets disjoint, then verify the integrated result against the original objective and current structured evidence.
