usensedata -query-china-company

usensedata-mcp-server-query-china-company exposes a set of MCP tools for querying Chinese company information. The server acts as an enterprise data service that leverages Usense data APIs to fetch company details, investment information, legal records and more, by taking structured tool calls from agents using the MCP protocol. The included tools cover verification of corporate names and leaders, fuzzy searches by keyword, and retrieval of core company attributes (basic info, overseas investments, change records, software copyrights, trademarks, court litigation, and abnormal operations). To use the server, specify the target tool and provide the required parameters. The server will orchestrate calls to the underlying data sources and return a consolidated response suitable for agent-driven conversations. Tools are designed to be composable; for example you can fuzzy search a company, then fetch its basic info and investment data in sequence to build a complete profile.

Common workflow examples:

• verify_company_name_and_president: provide operName (person) and entName (full company name) to confirm alignment.
• fuzzy_query_company: provide keyWord to retrieve a list of candidate company names matching the keyword, then select the correct full name for further queries.
• query_company_basic_info: provide entname to obtain core attributes like legal representative, registered capital, and unified social credit code.
• query_company_overseas_investments / query_company_change_records / query_company_software_copyright_info / query_company_trademark_list / query_company_court_litigation_related_info / query_company_abnormal_business_operation: provide the required entName (or name for some tools) to fetch respective records.

When integrating with an agent, you can chain these tools to build a complete data portrait for a given company (e.g., fuzzy_query_company → query_company_overseas_investments → query_company_basic_info) and present the results in a structured format to the user.

Prerequisites:

• Node.js and npm installed on your system
• Internet access to fetch packages

Installation steps (macOS/Linux):

# 1) Verify Node.js and npm are installed
node -v
npm -v

# 2) Install dependencies
npm install

# 3) Build TypeScript (if the project uses a build step)
npm run build

Installation steps (Windows):

# 1) Install Node.js from the official website and ensure npm is installed
node -v
npm -v

# 2) Install dependencies
npm install

# 3) Build (if applicable)
npm run build

Configure MCP server (example for macOS/Linux):

"mcpServers": {
  "usense-corp": {
    "command": "npx",
    "args": [
      "-y",
      "usensedata-mcp-server-query-china-company"
    ],
    "env": {
      "USENSEDATA_API_KEY": "your_api_key",
      "USENSEDATA_API_USERID": "your_api_userid"
    }
  }
}

Configure MCP server (Windows):

"mcpServers": {
  "usense-corp": {
    "command": "cmd",
    "args": [
      "/c",
      "npx",
      "-y",
      "usensedata-mcp-server-query-china-company"
    ],
    "env": {
      "USENSEDATA_API_KEY": "your_api_key",
      "USENSEDATA_API_USERID": "your_api_userid"
    }
  }
}

Notes:

• Ensure your API key and user ID are kept secret and not checked into source control.
• If you modify the package name or server id, update the mcp_servers configuration accordingly.

Tips and common issues:

• Environment variables USENSEDATA_API_KEY and USENSEDATA_API_USERID are required for authenticated requests. Set them in your MCP config or environment where the server runs.
• If you encounter network errors, verify that the target API endpoints are reachable from your deployment environment and that any necessary firewall rules allow outbound access.
• When upgrading the MCP server package, increment the version in package.json and publish to your npm registry if you maintain one.
• For Windows users, the command entry point uses a CMD wrapper to execute npx with the correct arguments; ensure cmd.exe is available in the environment path.
• The server assumes the base data source has consistent naming; if fuzzy_query_company returns multiple candidates, you may need to disambiguate using entName in subsequent calls.
• The provided example config uses the “usense-corp” server name; you can rename to fit your deployment naming conventions, but ensure references align across your orchestrating agent.
• The npm package name for this MCP server is usensedata-mcp-server-query-china-company.