Getting Started with SOLVE-IT MCP Server¶

This guide will get you up and running with the SOLVE-IT MCP Server in less than 5 minutes.

What You'll Need¶

• Docker installed (recommended) OR Python 3.11/3.12
• Basic understanding of Model Context Protocol (MCP)
• 5 minutes of your time

Choose Your Path¶

# Pull the latest image
docker pull 3soos3/solve-it-mcp:latest

# Run in HTTP mode
docker run -p 8000:8000 \
  -e MCP_TRANSPORT=http \
  -e HTTP_PORT=8000 \
  3soos3/solve-it-mcp:latest

# Test health endpoint
curl http://localhost:8000/healthz

# List available tools
curl -X POST http://localhost:8000/mcp/v1/messages \
  -H "Content-Type: application/json" \
  -d '{"method":"tools/list"}'

# Clone the knowledge base
git clone https://github.com/SOLVE-IT-DF/solve-it.git
cd solve-it
export SOLVE_IT_DATA_PATH=$(pwd)/data

# Clone this repository
git clone https://github.com/3soos3/solve-it-mcp.git
cd solve-it-mcp

# Install dependencies
pip install -r requirements.txt

# Run in stdio mode (for desktop MCP clients)
python3 src/server.py

# Or run in HTTP mode
python3 src/server.py --transport http --port 8000

# Test health endpoint
curl http://localhost:8000/healthz

INFO: SOLVE-IT MCP Server initialized successfully
INFO: Transport mode: stdio

{
  "mcpServers": {
    "solveit": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e", "MCP_TRANSPORT=stdio",
        "3soos3/solve-it-mcp:latest"
      ]
    }
  }
}

{
  "mcpServers": {
    "solveit": {
      "command": "python3",
      "args": ["/path/to/solve-it-mcp/src/server.py"],
      "cwd": "/path/to/solve-it-mcp",
      "env": {
        "SOLVE_IT_DATA_PATH": "/path/to/solve-it/data"
      }
    }
  }
}

Your First Query¶

Once the server is running, try these example queries:

Example 1: Search for Techniques¶

curl -X POST http://localhost:8000/mcp/v1/messages \
  -H "Content-Type: application/json" \
  -d '{
    "method": "tools/call",
    "params": {
      "name": "search",
      "arguments": {
        "keywords": "network analysis",
        "item_types": ["techniques"]
      }
    }
  }'

Example 2: Get Technique Details¶

curl -X POST http://localhost:8000/mcp/v1/messages \
  -H "Content-Type: application/json" \
  -d '{
    "method": "tools/call",
    "params": {
      "name": "get_technique_details",
      "arguments": {
        "technique_id": "T1001"
      }
    }
  }'

Example 3: Explore Relationships¶

curl -X POST http://localhost:8000/mcp/v1/messages \
  -H "Content-Type: application/json" \
  -d '{
    "method": "tools/call",
    "params": {
      "name": "get_weaknesses_for_technique",
      "arguments": {
        "technique_id": "T1001"
      }
    }
  }'

Understanding the Response¶

Successful responses will include:

• Content: The requested data (technique details, search results, etc.)
• Metadata: Information about the query and results
• Structure: JSON-formatted for easy parsing

Example response structure:

{
  "content": [
    {
      "type": "text",
      "text": "{ \"results\": [...] }"
    }
  ],
  "isError": false
}

Common Configuration Options¶

Environment Variables¶

Docker Compose Example¶

For persistent setups, use Docker Compose:

version: '3.8'
services:
  solve-it-mcp:
    image: 3soos3/solve-it-mcp:latest
    ports:
      - "8000:8000"
    environment:
      - MCP_TRANSPORT=http
      - HTTP_PORT=8000
      - LOG_LEVEL=INFO
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/healthz"]
      interval: 30s
      timeout: 3s
      retries: 3
    restart: unless-stopped

Save as docker-compose.yml and run:

docker-compose up -d

Next Steps¶

For Forensic Analysts¶

Learn how to use SOLVE-IT MCP in real investigations:

• Forensic Analysts Guide
• Tools Reference
• Common Investigation Patterns

For Researchers¶

Understand how to use SOLVE-IT MCP in your research:

• Researchers Guide
• Citation Information
• Tools Overview

For Developers¶

Set up a development environment:

• Local Development Guide
• Testing Guide
• Architecture Overview

For Production Deployments¶

Deploy to Kubernetes or production environments:

• Kubernetes Deployment
• Security Considerations
• Monitoring & Observability

Troubleshooting¶

Server Won't Start¶

Problem: Connection refused or server doesn't start

Solutions: - Verify Docker is running: docker ps - Check port availability: lsof -i :8000 - Review logs: docker logs <container-id>

SOLVE-IT Data Not Found¶

Problem: Failed to load SOLVE-IT knowledge base

Solutions: - Verify SOLVE_IT_DATA_PATH environment variable - Check that the path contains techniques.json, weaknesses.json, etc. - For Docker: the data is embedded in the image (no action needed)

Tools Not Appearing¶

Problem: MCP client doesn't show SOLVE-IT tools

Solutions: - Restart your MCP client - Check client configuration syntax - Verify server is running: curl http://localhost:8000/healthz - Check client logs for errors

For more help, see the Troubleshooting Guide.

Getting Help¶

• Documentation Issues: Check Troubleshooting
• Bug Reports: GitHub Issues
• Questions: GitHub Discussions
• Security Issues: See SECURITY.md