my skill

v1.0.0

行业深度分析报告生成器。 当用户提到行业分析、行业报告、市场分析、竞争格局分析、行业调研、industry analysis、 market research、分析XX行业、XX行业怎么样、XX行业值得进入吗时触发。 也适用于用户想了解某行业的市场规模、竞争态势、发展趋势、商业模式、投资机会, 或任何涉及"用结构...

0· 34·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for yangmanqi2104201431-ship-it/industry-analysis.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "my skill" (yangmanqi2104201431-ship-it/industry-analysis) from ClawHub.
Skill page: https://clawhub.ai/yangmanqi2104201431-ship-it/industry-analysis
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install industry-analysis

ClawHub CLI

Package manager switcher

npx clawhub@latest install industry-analysis
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The skill's name and description (industry analysis report generator) align with the instructions: perform web searches, collect sources, fill a provided template, render HTML/PDF and return the report. Required resources (web search via autoglm-websearch, reading the included template) match the stated purpose; there are no unexpected credential or binary requests.
Instruction Scope
SKILL.md stays on task: it mandates web search for sourcing, strict source/footnote rules, reading the bundled report template, producing Markdown → HTML → PDF/HTML output, and saving to the working directory. It does not instruct reading unrelated system files, environment variables, or sending data to third-party endpoints beyond web searching. It does call out using the autoglm-websearch skill for search, which is consistent with its purpose.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest install risk. It suggests tools (wkhtmltopdf / weasyprint) for PDF conversion but does not attempt to install them itself.
Credentials
The skill requests no environment variables, credentials, or config paths. The only external dependency is invoking autoglm-websearch (an external skill) for online searches; that dependency is proportional to the stated purpose but the trustworthiness/permissions of autoglm-websearch are outside this skill's scope.
Persistence & Privilege
always is false and the skill does not request permanent or elevated presence. It writes output to the current working directory (expected for a report generator) and does not modify other skills or system-wide configs.
Assessment
This skill is internally consistent: it will perform web searches (via the autoglm-websearch skill), collect and cite URLs, generate a Markdown report from the included template, render HTML and attempt PDF conversion, then save the file to the current working directory. Before installing, consider: (1) autoglm-websearch will perform network queries — verify you trust that search skill and its access permissions; (2) PDF conversion tools (wkhtmltopdf/weasyprint) may be required by your environment if you want PDF output; absence of an install step means the agent may fail PDF conversion if those tools aren't present; (3) the skill will write files to the agent’s working directory — ensure you are comfortable with that filesystem access; (4) the skill enforces strict non‑fabrication rules, but generated content may still include inferred data labeled as “[推断]” — review those carefully before taking operational decisions. No environment variables or credentials are requested by this skill itself.

Like a lobster shell, security has layers — review code before you run it.

latestvk972pdj2vs2bhdzbxpa7sep4g185nb67
34downloads
0stars
1versions
Updated 23h ago
v1.0.0
MIT-0

行业分析报告生成 Skill

工作流概览

1. 确认行业名称与分析范围
2. Web Search 分批搜索 → 采集数据
3. 读取 references/report-template.md,按模板填充生成 Markdown 报告
4. 将 Markdown 转换为精美 HTML(AI 自行设计样式与布局)
5. 输出文件 → 优先 PDF,降级为 HTML

🔴 质量红线(NEVER)

  1. NEVER 编造数据来源或 URL — 延伸阅读和参考文献的链接必须来自真实搜索结果
  2. NEVER 预测数据不加标注 — 所有预测值必须标注「E(预测)」
  3. NEVER 只呈现乐观分析 — 行业分析必须有风险维度和反面证据
  4. NEVER 忽略数据冲突 — 不同来源数据矛盾时,必须呈现分歧并标注可信度差异
  5. NEVER 掩盖数据缺失 — 无法获取可靠数据时,用 [推断] 标明,宁空勿编
  6. NEVER 所有玩家都给正面评价 — 至少指出每家企业的核心弱点或风险
  7. NEVER 用超过 18 个月的数据作为「最新数据」不加说明
  8. NEVER 忽略地域差异 — 全球性行业必须区分中国市场与海外市场

分析思维框架

写任何模块前,先问自己:

  • 时效性:最新数据是什么时候的?数据截止日期是否在报告顶部明确标注?
  • 立场平衡:是否只呈现了乐观一面?反面证据在哪里?
  • 数据密度:这一节有具体数字支撑,还是纯定性描述空话?
  • 行动导向:读者读完这一节能做出什么决策?还是看完等于没看?

数据采集策略

搜索主题优先级(按行业通用性排列)

  1. 行业发展历程 → 关键事件、政策节点、技术突破、并购重组
  2. 市场规模与增长 → 中英文各搜 1-2 个关键词组合
  3. 竞争格局与头部玩家 → 市场份额、排名
  4. 政策与监管动态 → 最新 2 年政策变化
  5. 技术趋势与创新 → 技术突破、专利趋势(必须是有足够代表性的技术)
  6. 投融资与资本动向 → 融资事件、上市动态
  7. 行业痛点与挑战 → 发展瓶颈、共性问题
  8. 深度报告与延伸阅读 → 权威咨询/学术/媒体

搜索执行原则

  • 使用 autoglm-websearch 技能进行搜索
  • 每个主题至少搜索 1 次中英文关键词组合
  • 搜索结果优先选择:政府/监管机构 > 权威咨询(麦肯锡/BCG/IDC/Gartner) > 行业协会 > 财经媒体 > 其他
  • 搜集过程中实时记录来源 URL、机构名、发布日期,后续用于脚注

降级策略

场景处理方式
Web Search 完全不可用报告顶部添加醒目免责声明,所有数据标注 [推断],注明"未联网验证"
部分主题搜索无结果对应模块标注 [推断],其余模块正常引用真实来源
搜索结果质量低(全是新闻碎片)扩大关键词范围,增加英文搜索词,尝试搜索权威机构官网
数据严重不足(<3 个可靠来源)缩减报告模块数量,聚焦已有数据的模块,其余标注"数据不足暂略"

HTML 渲染要求

将 Markdown 报告转为 HTML 时,AI 自行设计全部样式与布局。设计要求:

  • 封面:渐变背景 + 行业名称 + 报告日期 + 数据截止日期
  • 配色:审美有艺术感,避免"AI味"配色
  • 数据可视化:市场规模用横向条形图、五力用评分条、PEST 用四象限色卡、商业模式用九宫格、发展历程用时间线/里程碑图
  • 时间线设计:模块2(发展历程)用横向或纵向时间线,关键节点用卡片标注年份+事件+影响
  • 参考文献角标:正文中用圆形数字角标 [^N],悬停显示来源信息
  • 推断标注[推断] 用醒目黄色标签,报告顶部显示免责声明
  • 响应式 + 打印友好@media print 适配
  • 字体:中文 PingFang SC / Microsoft YaHei / Noto Sans SC 降级

不限制具体实现方式——用 Tailwind、手写 CSS、CSS Grid、Flexbox 都可以,以最终呈现效果为准。


References 加载指南

每次必读(生成报告前)

MANDATORY — 读取完整文件:生成 Markdown 报告前,必须读取 references/report-template.md 全文。 按模板中的 <!-- AI 规则 --> 注释和 {{占位符}} 填充内容。

Do NOT Load 其他文件——HTML 渲染由 AI 自行完成。


输出文件

  • 优先尝试 Python (wkhtmltopdf / weasyprint) 将 HTML 转 PDF
  • 失败时直接保存 HTML 文件
  • 保存到当前工作目录,文件名格式:industry_report_{{行业名}}_{{日期}}
  • 生成后通过消息发送给用户

注意事项

  1. Markdown 是中间产物:仅作为内容载体,最终交付物是 HTML 或 PDF
  2. 来源标注:所有明确数据用 [^N] 标注,参考文献条数与脚注完全对应
  3. 链接真实性:延伸阅读和参考文献的 URL 必须来自真实搜索结果

Comments

Loading comments...