菜单
Use when discovering user pain points from social media (Douyin), running pain point clustering analysis, or generating AI product solutions from user feedback
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
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/Leoyishou/personal-ai-company --skill deeppoint-ai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。
基于 SOC 职业分类
正在显示 SKILL.md
| name | deeppoint-ai |
| description | Use when discovering user pain points from social media (Douyin), running pain point clustering analysis, or generating AI product solutions from user feedback |
From social media data to actionable product insights. Collects Douyin comments by keyword, clusters them semantically (DBSCAN + embeddings), and generates AI product solutions.
Source: https://github.com/weiyf2/deeppoint-ai
| Item | Value |
|---|---|
| Repo | ~/usr/pac/product-bu/inbox/deeppoint-ai |
| Stack | Next.js 15 + Python 3.10+ + GLM-4 |
| Port | 3000 |
| API Key | GLM API from https://open.bigmodel.cn/ |
# 1. Clone
cd ~/usr/pac/product-bu/inbox
git clone https://github.com/weiyf2/deeppoint-ai.git
cd deeppoint-ai
# 2. Install dependencies
npm install
pip install -r requirements.txt
playwright install chromium
# 3. Configure .env.local
cat > .env.local << 'EOF'
GLM_API_KEY=<from secrets.env GLM_API_KEY>
GLM_MODEL_NAME=glm-4
GLM_EMBEDDING_MODEL=embedding-3
HEADLESS=true
EOF
# 4. Run
npm run dev
Load GLM key from ~/.claude/secrets.env if available (look for GLM_API_KEY).
| Tier | Data Points | Reliability |
|---|---|---|
| Exploratory | <50 | Low |
| Preliminary | 50-200 | Medium |
| Reliable | >=200 | High |
src/app/[locale]/ # i18n routes (zh/en)
src/components/ # UI components
lib/services/ # Job management, clustering, AI analysis
lib/crawlers/ # Douyin crawler (Playwright-based)
lib/semantic_clustering.py # Python DBSCAN clustering
lib/crawlers/douyin_new/) is NON-COMMERCIAL LICENSE only