Skip to main content
Glama
harishsg993010

Damn Vulnerable Model Context Protocol (DVMCP)

server_sse.py3.75 kB
from mcp.server.fastmcp import FastMCP, Context from mcp.server.sse import SseServerTransport from starlette.applications import Starlette from starlette.routing import Mount, Route from fastapi import FastAPI # Create a vulnerable MCP server for Challenge 6: Indirect Prompt Injection with SSE support class Challenge6Server: def __init__(self): self.name = "Challenge 6 - Indirect Prompt Injection" self.port = 9006 # Changed from 8006 to avoid conflicts self.mcp = FastMCP(self.name) self.app = FastAPI() # Add a document processing tool that is vulnerable to indirect prompt injection @self.mcp.tool() def process_document(document_text: str) -> str: """Process a document and extract key information Args: document_text: The text content of the document to process """ import os # Save the document for "processing" upload_dir = "/tmp/dvmcp_challenge6/user_uploads" os.makedirs(upload_dir, exist_ok=True) # Vulnerable: directly using user input in the processing logic # This allows for indirect prompt injection # Simulate document processing with a template processing_template = f""" DOCUMENT ANALYSIS SYSTEM ======================= PROCESSING DOCUMENT: {document_text} ANALYSIS RESULTS: The document appears to be a {len(document_text) > 1000 and 'long' or 'short'} text. It contains approximately {len(document_text.split())} words. KEY ENTITIES EXTRACTED: - {document_text.split()[:5] if document_text.split() else 'None found'} SENTIMENT: {'Positive' if 'good' in document_text.lower() else 'Negative' if 'bad' in document_text.lower() else 'Neutral'} END OF ANALYSIS """ return processing_template # Mount the SSE server self.mount_sse_server() def mount_sse_server(self): """Mount the SSE server to the FastAPI app""" self.app.mount("/", self.create_sse_server()) def create_sse_server(self): """Create a Starlette app that handles SSE connections and message handling""" transport = SseServerTransport("/messages/") # Define handler functions async def handle_sse(request): async with transport.connect_sse( request.scope, request.receive, request._send ) as streams: await self.mcp._mcp_server.run( streams[0], streams[1], self.mcp._mcp_server.create_initialization_options() ) # Create Starlette routes for SSE and message handling routes = [ Route("/sse", endpoint=handle_sse), Mount("/messages", app=transport.handle_post_message), ] # Create a Starlette app return Starlette(routes=routes) def run(self): """Run the server with uvicorn""" import uvicorn print(f"Starting {self.name} MCP Server") print("Connect to this server using an MCP client (e.g., Claude Desktop or Cursor)") print(f"Server running at http://localhost:{self.port}") print(f"SSE endpoint available at http://localhost:{self.port}/sse") uvicorn.run(self.app, host="0.0.0.0", port=self.port) # Run the server if __name__ == "__main__": server = Challenge6Server() server.run()

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/harishsg993010/damn-vulnerable-MCP-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server