← Back to Skills Marketplace
ling-qian

AI Market Research

by Chace · GitHub ↗ · v0.1.0 · MIT-0
cross-platform ⚠ suspicious
67
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install chace-ai-market-research
Description
整合 crawl4ai、trendradar、product-research 的全链路市场研究自动化技能
README (SKILL.md)

AI Market Research 组合技能

整合 crawl4ai、trendradar、product-research 的全链路市场研究自动化

定义

ai-market-research 是 OpenClaw 的组合技能,提供端到端的市场研究能力:

  1. crawl4ai - 深度网页抓取与结构化提取
  2. trendradar - 多平台热点监控与舆情分析
  3. product-research - 结构化市场分析框架(竞品/用户/趋势)
  4. agentmemory - 历史数据持久化与跨期对比

适用场景

  • 竞品进入新市场前的深度调研
  • 行业趋势追踪(每周/每月自动报告)
  • 新品发布前的市场环境扫描
  • 投资决策前的赛道分析

核心能力

输入参数

参数 类型 必填 说明
topic string 研究主题(产品名/行业/赛道)
depth enum 研究深度:quick(1小时)/standard(4小时)/deep(8小时+)
sources array 指定 crawl4ai 抓取的URL列表(不指定则自动发现)
output_format enum 输出格式:markdown/html/json(默认 markdown)
compare_previous bool 是否与历史研究对比(默认 true)

输出产物

  • 主报告 (research-report.md) - 结构化分析
  • 数据附件 (artifacts/) - 原始抓取内容、热度摘要、竞品表格
  • 记忆写入 - 自动保存到 agentmemory,支持长期追踪

工作流

graph LR
    A[任务: 研究 X] --> B[crawl4ai 深度抓取]
    A --> C[trendradar 热点聚合]
    B --> D[product-research 分析引擎]
    C --> D
    D --> E[生成结构化报告]
    E --> F[保存到 agentmemory]
    F --> G[推送结果]

阶段说明

  1. 数据采集 (Data Collection)

    • crawl4ai: 抓取产品官网、竞品页、评测文章、用户评论
    • trendradar: 获取微博/知乎/百度等平台的关联热点
  2. 数据分析 (Analysis)

    • product-research: 应用标准框架(SWOT/PEST/竞品矩阵)
    • AI 摘要: 提炼核心洞察、风险评估、机会点
  3. 知识存储 (Memory)

    • 将本次研究的关键结论存入向量数据库
    • 建立 topic → timestamp → findings 索引链
  4. 报告交付 (Delivery)

    • 生成 Markdown 报告(支持微信阅读)
    • 可选 HTML 可视化版本
    • 通过 OpenClaw 消息通道推送

使用示例

基本调用

# 快速研究(1小时)
ai-market-research --topic "宇树科技机器人" --depth quick

# 深度研究(8小时+)
ai-market-research --topic "人形机器人赛道" --depth deep --compare_previous true

指定数据源

ai-market-research \
  --topic "特斯拉FSD" \
  --sources "https://tesla.com/fsd" "https:// electrek.co/tesla-fsd-review" \
  --output_format markdown

配置依赖

MCP 服务

必须已启动:

  • crawl4ai (localhost:11235)
  • trendradar (localhost:3333)

技能启用

以下技能必须在 openclaw.json 中启用:

  • product-research
  • agentmemory (插件)
  • vector-memory(可选,用于历史对比)

环境变量

  • AI_API_KEY - 用于 trendradar AI 分析(如果启用)
  • OPENCLAW_WECHAT_TO - 微信推送目标(optional)

性能与成本

深度 预计耗时 crawl4ai 调用 trendradar 调用 LLM token 消耗
quick ~1h 5-10 URL 1次(当日数据) ~50k
standard ~4h 20-30 URL 3天历史 + 每日增量 ~200k
deep ~8h+ 50+ URL 7天历史 + 全平台 ~500k+

故障排除

常见问题

  1. crawl4ai JWT 错误 → 检查 CRAWL4AI_JWT 环境变量
  2. trendradar MCP 连接失败 → 确认 uv run python -m mcp_server.server 正在运行
  3. memory 写入失败 → 确认 agentmemory 插件已启用
  4. token 超限 → 降低 depth 或减少 sources 数量

安装

手动安装

# 克隆到技能目录
git clone https://github.com/yourusername/ai-market-research-skill.git \
  ~/.openclaw/workspace/.agents/skills/ai-market-research

# 启用技能(添加到 openclaw.json)
# 重启网关
openclaw gateway restart

通过 ClawHub(即将推出)

搜索 ai-market-research 并一键安装。

Roadmap

  • 真实 MCP 调用(替换模拟数据)
  • product-research 深度集成
  • 自动来源发现(Google 搜索 + 筛选)
  • 向量历史对比(vector-memory)
  • Webhook 推送自动化
  • Docker 容器化
  • 多语言报告支持

License

MIT © Chace

Usage Guidance
Install only if you are comfortable with research topics, generated findings, and source metadata being written locally and potentially saved to agentmemory. Do not enable channel/WeChat delivery or scheduled runs until you confirm the recipient, retention policy, and whether the current version is using simulated data or real integrations.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
The requested capabilities largely fit the stated market-research purpose: web/source collection, trend analysis, report generation, and historical comparison. However, the documentation overstates current live integrations while the roadmap and engine show crawl4ai, trendradar, and product-research are mostly simulated or planned.
Instruction Scope
The skill describes automatic historical comparison, memory writes, report saving, scheduled runs, and channel delivery, but does not clearly scope when these happen, how users disable them, or what data may be retained or sent.
Install Mechanism
Installation is manual via cloning into the OpenClaw skills directory and enabling it in config. The package references a bin/run wrapper, but the submitted artifact does not include that file, which is an integrity/operability mismatch rather than evidence of malware.
Credentials
External crawling/trend services and API keys are purpose-aligned, but the skill may process proprietary market topics, URLs, and competitive analysis without strong privacy, retention, or destination guidance.
Persistence & Privilege
engine.py writes reports automatically to a hard-coded local workspace path and unconditionally attempts agentmemory persistence after each run if available; README/Requirements also advertise auto-push to an OpenClaw/WeChat channel without clear opt-in safeguards.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install chace-ai-market-research
  3. After installation, invoke the skill by name or use /chace-ai-market-research
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
Initial release of ai-market-research – a full-stack market research automation skill. - Integrates crawl4ai (web crawling), trendradar (trend monitoring), and product-research (structured analysis) for end-to-end research workflows. - Supports depth control, custom sources, and multiple output formats. - Automatically stores findings with agentmemory and enables historical comparison. - Delivers structured reports and data artifacts for competitive, trend, and market analysis scenarios. - Includes setup requirements, troubleshooting, and performance guidance.
Metadata
Slug chace-ai-market-research
Version 0.1.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is AI Market Research?

整合 crawl4ai、trendradar、product-research 的全链路市场研究自动化技能. It is an AI Agent Skill for Claude Code / OpenClaw, with 67 downloads so far.

How do I install AI Market Research?

Run "/install chace-ai-market-research" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is AI Market Research free?

Yes, AI Market Research is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does AI Market Research support?

AI Market Research is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created AI Market Research?

It is built and maintained by Chace (@ling-qian); the current version is v0.1.0.

💬 Comments