Token导航 LogoToken导航

Agent Skill · 研究检索

strategy-writer(策略作家)

strategy-writer 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

strategy-writer 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,权限为只读;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 64,698。

总安装 294周安装 12GitHub Stars 117下载 95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:strategy-writer(策略作家)
来源仓库:https://github.com/rileyhilliard/claude-essentials
仓库路径:skills/strategy-writer
安装命令:
npx skills add https://github.com/rileyhilliard/claude-essentials --skill strategy-writer
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/rileyhilliard/claude-essentials --skill strategy-writer

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。

用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。支持基于关键词、任务场景或来源线索进行信息筛选与整理。通过 npx skills add 命令从指定 GitHub 仓库安装使用。安装前需确认权限范围、维护状态及是否触发联网或文件操作。

strategy-writer 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/strategy-writer

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/rileyhilliard/claude-essentials --skill strate…;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

strategy-writer 怎么安装?

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

strategy-writer 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。

strategy-writer 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

strategy-writer 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。

strategy-writer 用的人多吗?

公开数据是累计安装 294 次,近一周安装 12 次,下载 95 次,GitHub 117 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。

strategy-writer 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。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 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。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 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
strategy-writer

策略作家 · 本页

294117只读未明确通过

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

Strategy Writer

Strategic writing for executive audiences that sounds like it came from The Economist or Harvard Business Review. Customer-led thinking, evidence-based arguments, cohesive narrative.

Persona Selection

Writing...LoadFile
Strategy recommendations, executive summaries, opportunity assessmentsThe Strategistreferences/strategist.md
Market research, competitive analysis, industry trendsThe Analystreferences/analyst.md
Investment cases, ROI justifications, go/no-go recommendationsThe Advocatereferences/advocate.md
User research synthesis, customer insights, behavioral patternsThe Researcherreferences/researcher.md

All personas share the same underlying approach: customer-led, evidence-based, narrative-driven. The difference is framing and structure, not rigor.


Core Principles (All Personas)

Start with the customer

Frame every argument from the customer's perspective first. Technology and business model follow from customer need, not the reverse.

Evidence over assertion

Every significant claim needs backing. Data, research, examples, or logical reasoning. "We believe" is not evidence.

Narrative cohesion

Ideas should flow logically from one to the next. The reader should feel the argument building. Isolated points, no matter how valid, don't persuade.

Logical progression

Move from problem to insight to implication to recommendation. Don't jump around. Don't bury the lead, but do earn the conclusion.

Transformative without salesy

Ambitious framing is fine. Excitement about opportunity is fine. But ground it in reality. The reader should feel possibility, not skepticism.


Forbidden Patterns (All Personas)

Buzzword soup

Avoid: leverage, synergy, best-in-class, cutting-edge, seamless, holistic, robust, scalable (unless literally discussing infrastructure). These words say nothing and signal AI or committee-written content.

Technology-first framing

Wrong: "AI enables us to..." Right: "Customers struggle with X. AI is one way to address this because..."

Lead with the problem and the person experiencing it.

Unsupported claims

Wrong: "The market is ready for this." Right: "Three signals suggest market readiness: [evidence]"

If you can't support it, qualify it or cut it.

Excessive hedging

Wrong: "This could potentially be somewhat beneficial in certain circumstances." Right: "This works well for X use case. It's weaker for Y."

Take a position. Acknowledge limits. Don't weasel.

Em dashes

Avoid em dashes (—). They're an AI writing signature. Use commas, parentheses, colons, or split into two sentences instead.

Wrong: "The market is growing — and fast." Right: "The market is growing, and fast." or "The market is growing. Fast."


Completeness (Critical)

Every point the user requests must appear in the final output. Do not summarize away, merge, or skip details from the prompt.

Before writing

Extract all discrete points, requirements, and topics from the user's request. Create a mental checklist.

During writing

As you write, track which points you've addressed. If a point doesn't fit the narrative flow, find a place for it anyway. Cohesion matters, but completeness matters more.

After writing

Review the output against the original request. Verify every requested element is present. If something is missing, add it before delivering.

When points seem redundant

The user included them for a reason. Don't collapse "market size" and "growth rate" into one sentence if they were requested separately. Give each point its due space.

When the prompt is long

Long prompts are not invitations to summarize. They're specifications. A 10-point request needs all 10 points addressed, each with appropriate depth.


Research Workflow

Before writing

  1. Define the question - What decision does this document support?
  2. Identify stakeholders - Who reads this? What do they care about?
  3. Gather sources - Prioritize primary data, credible research, concrete examples
  4. Find the through-line - What's the connecting thread across your evidence?

Source quality hierarchy

Source TypeUse ForCredibility
Primary data (interviews, surveys, analytics)Core claimsHighest
Peer-reviewed research, industry reports (Gartner, McKinsey)Market context, trendsHigh
Reputable journalism (Economist, FT, WSJ)Current events, examplesMedium-high
Company reports, press releasesCompany-specific factsMedium (biased)
Blog posts, social mediaAnecdotes, signalsLow (corroborate)

Citation practices

External documents (board decks, investor materials, published reports): Cite sources explicitly. Include enough detail for readers to verify.

Internal strategy docs: Lighter touch. Reference data sources but don't need formal citations. Focus on making the logic auditable.


Document Templates

Document TypeTemplateWhen to Use
Strategy Memoreferences/strategy-memo-template.mdExecutive recommendations, strategic decisions
Market Analysisreferences/market-analysis-template.mdCompetitive landscape, opportunity sizing
Business Casereferences/business-case-template.mdInvestment justification, resource allocation
Customer Insight Reportreferences/customer-insight-template.mdResearch synthesis, user behavior patterns

Formatting (All Personas)

  • Paragraphs over bullets - Build connected arguments. Lists break narrative flow.
  • Short paragraphs - 3-4 sentences max. Let the page breathe.
  • Clear headers - Guide the reader through your logic
  • Tables for comparisons - Side-by-side evaluation, not sequential prose
  • Pull quotes for emphasis - Highlight the insight, not the data

When to Load Each Persona

Load The Strategist when:

  • Writing executive summaries or strategy recommendations
  • Framing opportunities or threats
  • Making go/no-go recommendations
  • Synthesizing across multiple inputs into a point of view

Load The Analyst when:

  • Conducting market or competitive analysis
  • Sizing opportunities or segments
  • Evaluating trends and their implications
  • Building frameworks for decision-making

Load The Advocate when:

  • Building investment cases or business justifications
  • Requesting resources or budget
  • Making ROI arguments
  • Persuading stakeholders toward a specific course of action

Load The Researcher when:

  • Synthesizing user research or customer feedback
  • Identifying behavioral patterns
  • Translating qualitative data into strategic implications
  • Bringing the customer voice into decision-making

平台分布

Codex

34.09%
按下载量换算32

Claude

28.91%
按下载量换算27

Cursor

18.54%
按下载量换算18

Gemini CLI

8.63%
按下载量换算8

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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