lex-intel 0.1.0

Chinese AI intelligence MCP server — daily curated signals from 11 Chinese-language sources, translated and scored for enterprise AI/tech readers

Lex Intel — Chinese AI Intelligence for AI Agents

MCP server that gives AI agents access to curated Chinese AI/tech intelligence. Daily signals from 11 Chinese-language sources (36Kr, Huxiu, CSDN, Caixin, Zhidx, and more), translated to English, scored by relevance, and served through semantic search.

When to Use This

• Your agent needs to answer "What's happening in Chinese AI?"
• You're researching Chinese tech companies, funding, regulation, or breakthroughs
• You want daily briefings on Chinese AI developments in Bloomberg-style format
• You need trend analysis: which categories of Chinese AI are rising or declining

Quick Start

pip install lex-intel

Add to your MCP client config:

{
  "mcpServers": {
    "lex-intel": {
      "command": "lex-intel",
      "env": {
        "SUPABASE_URL": "your-url",
        "SUPABASE_ANON_KEY": "your-key",
        "PINECONE_API_KEY": "your-key"
      }
    }
  }
}

Tools

Tool	When To Call It
lex_search_articles	Search for articles about a topic, company, or technology in Chinese AI
lex_get_briefing	Get the latest daily Bloomberg-style briefing (LEAD/PATTERNS/SIGNALS/WATCHLIST/DATA)
lex_get_signals	Find high-relevance (4-5) developments from recent days, grouped by category
lex_get_trending	See which Chinese AI categories have rising or declining coverage momentum
lex_list_sources	Check which sources are active and their signal quality
lex_get_article	Get full article details (body text, URL) after finding it via search

Data Sources

11 Chinese-language outlets scraped daily: 36Kr, Huxiu, InfoQ China, CSDN, SAP China, Kingdee, Yonyou, Leiphone, Caixin, Jiemian, Zhidx

Plus Gmail newsletters from Chinese tech sources.

Categories

Articles are classified into 13 categories: funding, m_and_a, investment, product, regulation, breakthrough, research, open_source, partnership, adoption, personnel, market, other

Each article is scored 1-5 for relevance to enterprise AI/tech readers.

License

MIT