https://github.com/wuye-ai/mcp-server-wuye-ai
CRIC物业AI MCP Server
https://github.com/wuye-ai/mcp-server-wuye-ai
agent ai mcp mcp-server property-management
Last synced: 14 days ago
JSON representation
CRIC物业AI MCP Server
- Host: GitHub
- URL: https://github.com/wuye-ai/mcp-server-wuye-ai
- Owner: wuye-ai
- Created: 2025-04-23T09:27:41.000Z (19 days ago)
- Default Branch: main
- Last Pushed: 2025-04-25T09:29:56.000Z (17 days ago)
- Last Synced: 2025-04-25T09:49:58.937Z (17 days ago)
- Topics: agent, ai, mcp, mcp-server, property-management
- Size: 467 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.en.md
Awesome Lists containing this project
- awesome-mcp-servers - CRIC Wuye AI - Interact with capabilities of the CRIC Wuye AI platform, an intelligent assistant specifically for the property management industry. (Official Servers)
- awesome-mcp-servers - Wuye-AI/mcp-server-wuye-ai - CRICプロパティAIプラットフォームに接続するMCPサーバーです。CRICプロパティAIは、克而瑞がプロパティ業界向けに開発したインテリジェントAIアシスタントです。 (サーバー実装 / 🔎 <a name="search"></a>検索)
README
# CRIC Wuye AI MCP Server
--------------
[](https://github.com/wuye-ai/mcp-server-wuye-ai/blob/master/README.md)
[](https://github.com/wuye-ai/mcp-server-wuye-ai/blob/master/README.en.md)
Available on: [**[NPM]**](https://www.npmjs.com/package/@wuye-ai/mcp-server-wuye-ai) [**[MCP.so]**](https://mcp.so/server/CRIC%E7%89%A9%E4%B8%9AAI/CRIC)
(More MCP platform integrations coming soon…)
--------------
## Overview
**CRIC Wuye AI** (CRIC物业AI) is an intelligent assistant developed by [CRIC](http://www.cricchina.com/) specifically for the property management industry. It was [officially launched](https://mp.weixin.qq.com/s/GC4V1M6N199Ay2f3kZan_Q) on April 25, 2025.
*P.S. Wuye (物业) means Property Management in Chinese.*
Leveraging a combination of industry knowledge base construction, multimodal large models, and RAG technology, **CRIC Wuye AI** integrates five core capability modules: **Industry Research**, **Laws & Regulations**, **Community Governance**, **Project Operations**, and **Content Writing**. It also expands into two smart agents focused on **News & Public Opinion** and **Talent Training** based on vertical industry knowledge.
## Core Capabilities
CRIC builds its competitive edge in the Wuye AI collaboration space through three main capabilities:
- **Data Asset Transformation:** Converts over 1 billion characters of industry text and terabytes of multimodal data into high-quality datasets tailored for the property sector, along with a custom industry data quality evaluation system to ensure accuracy and reliability.
- **Scenario Penetration:** Focused on 20+ vertical business scenarios in property management, leveraging targeted knowledge bases to enable precise matching.
- **Ecosystem Evolution:** Real-time monitoring of 500+ trusted information and data sources daily, processing over 100,000 real-time data updates. Achieves over 90% accuracy in areas like policy alerts, business opportunity mining, and tender analysis—creating a continuously evolving industry AI knowledge hub.
## MCP Server Features
**CRIC Wuye AI MCP Server** is a server-side implementation based on the [Model Context Protocol](https://modelcontextprotocol.github.io/), delivering part of the atomic capabilities of the **CRIC Wuye AI** platform. The current version provides the following tools:
- **Get Daily News:** Retrieve the property industry’s daily news report based on a `YYYY-MM-DD` date format.
- **Get Hot Topics List:** Retrieve categorized lists of trending questions in the property management sector within CRIC Wuye AI.
- **List Available Knowledge Bases:** List available knowledge bases for querying.
- **Search Knowledge Base:** Specify a knowledge base ID and query to retrieve matching results as document snippets or full text for AI reference.
More features are coming soon. Stay tuned.
## Obtaining an Access Token
To use the CRIC Wuye AI MCP Server, you must first obtain an **Access Token**. Please contact our support staff via email at [[email protected]](mailto:[email protected]).
## Getting Started
### 1. SSE Mode (http)
#### 1.1 Run
You can run your own MCP Server with HTTP mode enabled, or use the official URL we provide.
##### A. Self-hosted:
```bash
MODE=http PORT=3011 npx -y @wuye-ai/mcp-server-wuye-ai
```
After successful startup, your MCP Server URL is `http://localhost:3011/sse/mcp`.
##### B. Use Official:
Alternatively, you can use our official MCP Server URLs:
- Testing: `https://mcp.wuye-ai-staging.cricbigdata.com/sse/mcp`
- Production: `https://mcp.wuye-ai.cricbigdata.com/sse/mcp`
#### 1.2 Test
You can test the SSE MCP Server using MCP Inspector or third-party tools.
##### MCP Inspector:
Once the server is running, use [MCP Inspector](https://github.com/modelcontextprotocol/inspector) to test service status:
```bash
npx @modelcontextprotocol/inspector
```
After launching, open the Web UI in your browser (default: http://127.0.0.1:6274/) and configure the connection:
1. Set **Transport Type** to `SSE` and **URL** to the MCP Server URL obtained earlier.
2. Under **Authentication**, set **Header Name** to `Authorization` and **Bearer Token** to your **CRIC Wuye AI Access Token**.
3. Click **Connect**. Once connected, the left panel will show the connection status.
You can now use MCP Inspector to test CRIC Wuye AI MCP Server. See the [MCP Inspector Documentation](https://modelcontextprotocol.io/docs/tools/inspector) for more details.
##### Third-Party Tools:
When using SSE mode with other tools, set the `Authorization` header with your Access Token. For example, in [Cline](https://cline.bot/):
```json
{
"mcpServers": {
"CRIC-Wuye-AI": {
"transportType": "sse",
"url": "{{Your MCP Server URL}}",
"headers": {
"Authorization": "Bearer {{Your Access Token}}"
}
}
}
}
```
Note: Some tools using [@modelcontextprotocol/typescript-sdk](https://github.com/modelcontextprotocol/typescript-sdk) may not properly apply HTTP headers ([known issue](https://github.com/modelcontextprotocol/typescript-sdk/issues/317)). As a workaround, you can also pass the token via the query string:
```json
{
"mcpServers": {
"CRIC-Wuye-AI": {
"transportType": "sse",
"url": "{{Your MCP Server URL}}?token={{Your Access Token}}"
}
}
}
```
### 2. Stdio Mode
#### 2.1 Run
Stdio mode is also supported. To start the server:
```bash
CRIC_WUYE_AI_ACCESS_TOKEN={{Your Access Token}} npx -y @wuye-ai/mcp-server-wuye-ai
```
#### 2.2 Test
You can test using either third-party tools or MCP Inspector. Typically, Stdio mode does not require manual startup, as tools will start it automatically.
##### MCP Inspector:
To use Stdio in MCP Inspector:
1. Set **Transport Type** to `Stdio`, **Command** to `npx`, and **Arguments** to `-y @wuye-ai/mcp-server-wuye-ai`.
2. Under **Environment Variables**, set `CRIC_WUYE_AI_ACCESS_TOKEN` to your **Access Token**.
3. Click **Connect**. The Inspector will auto-run the server and connect.
You can now test CRIC Wuye AI MCP Server using MCP Inspector.
##### Third-Party Tools:
For Stdio access via third-party tools, specify the Access Token using the `CRIC_WUYE_AI_ACCESS_TOKEN` environment variable. For example, in Cline:
```json
{
"mcpServers": {
"CRIC-Wuye-AI": {
"transportType": "stdio",
"command": "npx",
"args": [
"-y",
"@wuye-ai/mcp-server-wuye-ai"
],
"env": {
"CRIC_WUYE_AI_ACCESS_TOKEN": "{{Your Access Token}}"
}
}
}
}
```
## Optional Configuration
You can configure how CRIC Wuye AI MCP Server runs via environment variables. Supported options:
| Parameter | Default | Description |
|----------------------------------|------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `MODE` | `stdio` | Run mode. Options: `stdio`, `http`. |
| `HOSTNAME` | `0.0.0.0` | Hostname for HTTP mode. `0.0.0.0` binds to all your IP addresses. |
| `PORT` | `3011` | Port for HTTP mode. |
| `CRIC_WUYE_AI_ACCESS_TOKEN` | *none* | CRIC Wuye AI Access Token. If not provided, will use the `Authorization` HTTP header value. |
| `CRIC_WUYE_AI_PROVIDER_API_BASE` | `https://export.wuye-ai.cricbigdata.com` | Backend API base URL (***this is not the MCP Server URL***). Optional values: `https://export.wuye-ai-staging.cricbigdata.com` (testing) or `https://export.wuye-ai.cricbigdata.com` (production). |