Token导航 LogoToken导航

Agent Skill · 研究检索

orchestrate(orchestrate 搜索)

用于辅助文档、README、Markdown、说明文和内容稿件的整理与改写。它适合让 Agent 提炼结构、补齐章节、统一术语、检查链接或把零散材料整理成可读文档。使用时应保留项目已有事实、命令和路径,不要把未确认的信息写成确定结论;涉及对外文案时,还需要控制语气,避免过度营销或夸大能力。

orchestrate 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为执行命令;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 62,244。

总安装 339周安装 14GitHub Stars 140下载 111核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:orchestrate(orchestrate 搜索)
来源仓库:https://github.com/haowjy/creative-writing-skills
仓库路径:skills/orchestrate
安装命令:
npx skills add https://github.com/haowjy/creative-writing-skills --skill orchestrate
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/haowjy/creative-writing-skills --skill orchestrate

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

orchestrate 用于辅助文档、README 和内容稿件的整理与改写,适合提炼结构、统一术语或检查链接。适用于需要梳理 Markdown、说明文或对外文案可读性的场景,帮助 Agent 组织零散材料成规范文档。使用时保留项目已有事实和命令,避免写成确定结论;涉及对外文案时需控制语气,防止过度营销。

通过 GitHub 安装,支持 Codex、Claude、Cursor 和 Gemini CLI,需确认权限范围和维护状态。建议结合原始 README 核验具体用法,尤其注意是否会触发联网或文件读写操作。

来源仓库内路径:skills/orchestrate

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/haowjy/creative-writing-skills --skill orchest…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要执行命令,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

orchestrate 怎么安装?

在终端执行 npx skills add https://github.com/haowjy/creative-writing-skills --skill orchestrate。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

orchestrate 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

orchestrate 需要什么权限?

权限级别是执行命令。安装流程涉及命令执行,可能通过 npx skills add https://github.com/haowjy/creative-writing-skills --skill orchestrate 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

orchestrate 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

orchestrate 用的人多吗?

公开数据是累计安装 339 次,近一周安装 14 次,下载 111 次,GitHub 140 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

orchestrate 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要执行命令,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 不能接受 Skill 在你机器上执行命令的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
orchestrate

orchestrate 搜索 · 本页

339140执行命令未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Orchestrate

Orchestrators coordinate — they don't produce content. Your value is the vantage point: seeing across phases, catching drift between what the author intended and what agents produced, routing corrections. When you drop into writing prose, editing files, or producing content yourself, you lose that view.

Delegation Discipline

Delegate everything substantive. Every creative action is a spawn. You evaluate results, you don't produce them. Writing your own drafts bypasses the critique and revision lanes that catch what the writer can't see in their own work — even for "small fixes."

Use meridian spawn, not built-in Agent tools. Spawns persist reports, enable model routing across providers, and are inspectable after the session ends. Built-in agent tools lack these properties.

Match prompt scope to agent scope. Before spawning, consider what the agent is built to do. A writer expects a scene brief with style files. A critic expects a focus area and the draft plus reference files. A brainstormer expects a scoped question with constraints. Shape the prompt to fit the agent's role — prompts shaped for the wrong role invite scope creep or misplaced effort.

Convergence

Loops run until convergence, not a fixed count. Convergence = no new substantive findings.

Substantive = would change how the scene reads, catch a continuity error, or damage the reader's experience. Non-substantive = style preferences, minor wording, things the author already decided.

When critics disagree or go in circles, you have context they don't — the full brief, prior iterations, the author's intent. Make the call, note the reasoning in your report so the author can understand what was decided and why.

Two to three revision cycles is typical. Past four without convergence usually signals a structural problem in the brief or outline — report this rather than continuing to iterate on symptoms.

Critique Synthesis

Synthesize at the pattern level rather than forwarding individual critic reports. "3 of 4 critics flagged pacing in scenes 2-4" is more actionable than four separate reports.

When triaging findings, categorize against the four reader reward channels (documented in the writing-principles skill):

  • Transportation — narrative coherence, POV consistency, continuity, middle-passage drift
  • Aesthetic — voice, rhythm, syntactic variety, generic vs specific detail
  • Social simulation — character motivation, emotional access through behavior not labels, ambiguity preservation
  • Flow — pacing, over-elaboration, exposition placement, challenge-fit

"Three critics flagged findings that damage transportation" is more actionable triage than a raw list. The author can decide whether transportation problems or aesthetic problems are more urgent given where the draft stands. Prioritize findings by impact on the reader's experience.

Artifacts Are State

Anything not on disk doesn't survive your spawn ending. Conversation context compacts; session history gets summarized; your reasoning evaporates. The artifacts are the resume contract — if another agent picks up where you left off, the files are all they have.

Keep artifacts current as you go. Promote durable knowledge to the kb before closing out a work item. Decision reasoning, author preferences, rejected alternatives — if a future agent would need it, write it down now.

平台分布

Codex

31.65%
按下载量换算35

Claude

30.76%
按下载量换算34

Cursor

19.61%
按下载量换算22

Gemini CLI

10.15%
按下载量换算11

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/haowjy/creative-writing-skills --skill orchestrate 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills