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/imbue-ai/catalyst --skill streamline-theory-variations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
基于 SOC 职业分类
| name | streamline-theory-variations |
| description | Streamline a theory down to its core essence, selecting a few different options. |
| argument-hint | theory ID (e.g. T_20260414_143100_d4e5f6) |
Arguments: $ARGUMENTS
The arguments contain a theory ID (like T_20260414_...).
You will run the streamline-theory skill multiple times, each in a separate parallel subagent, and each with a different instruction for which key story to focus on.
In particular:
streamline-theory with the instruction to focus on "the most novel aspect". Also pass it the input theory ID.streamline-theory with the instruction to focus on "the most insightful aspect". Also pass it the input theory ID.streamline-theory with the instruction to focus on "the most foundational aspect". Also pass it the input theory ID.streamline-theory with the instruction to focus on "the most universally applicable aspect". Also pass it the input theory ID.streamline-theory with the instruction to focus on "the most speculative and ambitious aspect". Also pass it the input theory ID.All subagents can run in parallel. Each one will report back a new theory ID (e.g. T_20260414_150000_x1y2z3). Wait for all subagents to finish and collect their final result messages containing the returned theory IDs. Report the list of returned theory IDs as the output of this skill.