Three ways to give your AI assistant direct access to OpenBook's real-world data
For Claude Desktop, Cursor, Cherry Studio, Windsurf, and all MCP clients
{
"mcpServers": {
"openbook": {
"url": "https://openbook.now/api/mcp"
}
}
}search_signalsSearch signals by city, category, type, keyword
params: city, category, signal_type, keyword
get_signal_detailGet full details of a single signal
params: signal_id
submit_signalSubmit a new signal with structured data
params: city, category, title, content, ...
submit_signal_textSubmit signal via natural language (AI auto-extracts)
params: text
list_citiesList covered cities and signal counts
params: none
For WeChat AI assistants: WorkBuddy, OpenClaw, QClaw, KimiClaw
Open WorkBuddy → Skills → Import Skill → Select the unzipped openbook folder
Or manually copy to: C:\Users\YourName\.workbuddy\skills\
Copy the openbook folder to ~/.openclaw/skills/
Type @openbook followed by your question, AI will automatically query OpenBook:
For developers integrating into GPT Actions, Dify, Coze, etc.
GET https://openbook.now/api/trpc/signals.smartSearch?input={"json":{"query":"北京隔音好的小区"}}Accepts natural language queries, auto-detects city, category, and keywords
GET https://openbook.now/api/trpc/signals.query?input={"json":{"city":"Beijing","category":"housing","keyword":"soundproofing"}}Optional params: city, category (housing/food/jobs), signal_type (recommendation/warning/neutral/tip), keyword
{
"result": {
"data": {
"json": {
"signals": [
{
"id": "bj-housing-001",
"title": "万柳书院 隔音评价",
"city": "Beijing",
"category": "housing",
"signal_type": "recommendation",
"content": "双层楼板+专业隔音垫，隔音极佳",
"tags": ["soundproofing", "haidian"],
"date": "2025-03-15"
}
],
"total": 42
}
}
}
}