Skip to content

QUICKSTART.md

Latest commit

 

History

History
222 lines (174 loc) · 4.33 KB

QUICKSTART.md

File metadata and controls

222 lines (174 loc) · 4.33 KB

GhidraInsight Quick Start Guide

Installation (5 minutes)

Option 1: Docker (Recommended)

git clone https://github.com/yourusername/GhidraInsight.git
cd GhidraInsight
docker-compose up

Access dashboard: http://localhost:3000

Option 2: Manual Setup

1. Java Ghidra Plugin

cd ghidra-plugin
./gradlew build
# Copy JAR to Ghidra extensions directory
cp build/libs/GhidraInsight-1.0.0.jar $GHIDRA_INSTALL_DIR/Extensions/Ghidra/

2. Python MCP Server

cd python-mcp
pip install -e .
ghidrainsight-server --host 0.0.0.0 --port 8000

3. Web Dashboard

cd web-dashboard
npm install
npm run dev

Basic Usage

1. Upload Binary

  1. Open http://localhost:3000
  2. Drag & drop your binary or click to select

2. Run Analysis

  1. Click "Start Analysis"
  2. Select features:
    • ✅ Crypto Detection
    • ✅ Vulnerability Analysis
    • ✅ Taint Analysis
  3. Wait for results (typically 30-60 seconds)

3. Review Results

  • Crypto: Detected algorithms and their locations
  • Vulnerabilities: CVSS scores and remediation hints
  • Taint: Data flow paths from sources to sinks

4. Ask AI Questions

Use the chat interface to ask:

  • "What functions use crypto?"
  • "How does user input flow to system calls?"
  • "What are the main security risks?"

Command Line Usage

Analyze a binary

ghidrainsight analyze --file binary.elf \
  --features crypto,taint,vulnerabilities \
  --output report.json

Start server with custom config

ghidrainsight-server --config config.yaml \
  --host 0.0.0.0 \
  --port 8000 \
  --log-level DEBUG

Query the API

curl -X POST http://localhost:8000/api/analyze \
  -F "file=@binary.elf" \
  -H "X-API-Key: your-api-key" \
  -d "features=crypto,vulnerabilities"

Python Integration

import asyncio
from ghidrainsight import GhidraInsightClient

async def main():
    async with GhidraInsightClient("http://localhost:8000") as client:
        # Analyze binary
        results = await client.analyze_binary(
            "binary.elf",
            features=["crypto", "vulnerabilities"]
        )
        print(f"Found {len(results['vulnerabilities'])} issues")
        
        # Analyze specific function
        function = await client.analyze_function("0x401234")
        print(function)

asyncio.run(main())

AI Integration (ChatGPT/Claude)

Using GhidraInsight with ChatGPT

import openai

response = openai.ChatCompletion.create(
    model="gpt-4",
    tools=[{
        "type": "function",
        "function": {
            "name": "analyze_binary",
            "description": "Analyze a binary with GhidraInsight",
            "parameters": {
                "type": "object",
                "properties": {
                    "file_path": {"type": "string"}
                },
                "required": ["file_path"]
            }
        }
    }],
    messages=[
        {"role": "user", "content": "Analyze this binary: binary.elf"}
    ]
)

Configuration

Environment Variables

export GHIDRA_INSTALL_DIR=/path/to/ghidra
export JWT_SECRET=your-secret-key-here
export API_KEY=your-api-key-here
export LOG_LEVEL=INFO
export RATE_LIMIT=60

Config File (YAML)

server:
  host: 0.0.0.0
  port: 8000
  
auth:
  enabled: true
  provider: jwt
  secret: ${JWT_SECRET}
  
security:
  rate_limit:
    requests_per_minute: 60
  cors:
    allowed_origins:
      - http://localhost:3000

Troubleshooting

Port Already in Use

# Find process using port 8000
lsof -i :8000

# Kill the process
kill -9 <PID>

# Or use different port
ghidrainsight-server --port 8001

Out of Memory

# Increase JVM memory for Java plugin
export JAVA_OPTS="-Xmx8g"

WebSocket Connection Failed

  • Ensure firewall allows port 8001
  • Check CORS configuration
  • Verify server is running

High Memory Usage

  • Close unused analysis tabs
  • Reduce file size or split analysis
  • Check for memory leaks with jmap

Next Steps

Happy analyzing! 🔍