Sample-Related Foundational Data

About XLab

XLab is a research team under QiAnXin focused on large-scale network security, threat analysis and attribution, and the development of multidimensional security data platforms.

As one of the earliest teams in China to conduct security research and threat intelligence production using large-scale data, XLab has built the country’s first PassiveDNS system. It has also developed several leading foundational data systems, including those for Netflow, Whois, digital certificates, IP addresses, and malicious samples.

Sample-Related Foundational Data

These tools are used to query sample-related information within the foundational network data set. Currently, this includes access to sandbox analysis data.

Tools

1. sandbox
   • Allows queries using domain names, IP addresses, or sample SHA1 hashes.
   • Returns a summary of the sample’s network behavior.

Installation & Usage

Notes

This toolset is currently in its trial version for basic security data queries. Under high-concurrency conditions, the backend may throttle network access.

For high-frequency or production-level use cases, please switch to the official version.

MCP Service Endpoint

The endpoint for QiAnXin XLab's remote MCP service is: https://fdp.qianxin.com/mcp/v1/sample/mcp/

Official Version Installation Guide

1. The official version supports streamable-http access.
2. When using the tool, you must include two HTTP headers: fdp-access and fdp-secret.
   • To obtain these credentials, please contact QiAnXin XLab.
3. When writing agent code to connect with the MCP service, include the fdp-access and fdp-secret headers in your HTTP requests. Here's an example using smolagents:

   from smolagents import ToolCollection
   from smolagents.agents import ToolCallingAgent
   from smolagents.models import OpenAIServerModel
   
   def main():
   	with ToolCollection.from_mcp(
   		{
   			"url": "https://fdp.qianxin.com/mcp/v1/sample/mcp/",
   			"transport": "streamable-http",
   			"headers": {
   				"fdp-access": "xxxx",
   				"fdp-secret": "yyyy",
   			},
   		},
   		trust_remote_code=True,
   	) as tools:
   		agent = ToolCallingAgent(
   			tools=[*tools.tools],
   			model=OpenAIServerModel(
   				model_id="YOUR-LLM-MODEL-ID",
   				api_base="YOUR-LLM-MODEL-API-URL",
   				api_key="YOUR-LLM-MODEL-API-KEY",
   			),
   		)
   		agent.run("Query sample-related data associated with the domain www.example.com.")
   
   if __name__ == "__main__":
   	main()
   

4. If you are using Claude Desktop, you can leverage the mcp-remote library for proxy forwarding. During the forwarding process, configure the necessary HTTP headers as shown below:

   {
   	"mcpServers": {
   		"fdp_ip": {
   			"command": "npx",
   			"args": [
   				"-y",
   				"mcp-remote@latest",
   				"https://fdp.qianxin.com/mcp/v1/sample/mcp/",
   				"--header",
   				"fdp-access:xxxx",
   				"--header",
   				"fdp-secret:yyyy"
   			]
   		}
   	}
   }