AI-powered Model Context Protocol server for CTF challenge solving and automation.
🎯 CTF Categories Supported:
- Web Exploitation
- Cryptography
- Binary Exploitation (PWN)
- Forensics
- Reverse Engineering
- OSINT
- Miscellaneous
🤖 AI-Powered Tools:
- Automated workflow generation
- Smart tool selection
- Challenge solving assistance
- Exploit template generation
- Install dependencies:
pip install mcp requests- Start the Glass CTF server:
cd C:\MY_PERSONAL_SPACE\Glass
npm start- Configure the MCP in your AI client using
glass-ctf-mcp.json
create_ctf_workflow- Create challenge workflow with recommended toolsexecute_ctf_challenge- Execute workflow against target
analyze_binary- Analyze binary security featuresgenerate_exploit- Generate pwntools exploit template
solve_crypto- Auto-solve crypto challenges
analyze_forensics_file- Analyze files for hidden data
analyze_reverse_binary- Analyze binaries for reversing
scan_web_challenge- Scan web applications
execute_command- Execute shell commandsserver_health- Check server status
# Create a PWN workflow
create_ctf_workflow(category="pwn", difficulty="medium")
# Analyze a binary
analyze_binary(binary_path="/path/to/binary")
# Generate exploit
generate_exploit(binary_path="/path/to/binary", host="ctf.example.com", port=9999)
# Solve crypto
solve_crypto(ciphertext="SGVsbG8gV29ybGQh", method="auto")Update glass-ctf-mcp.json to change:
- Server URL (default: http://127.0.0.1:8888)
- Timeout settings
- Auto-allow permissions
The MCP server integrates with AI assistants (Claude, etc.) to provide CTF automation capabilities through natural language.
Example queries:
- "Analyze this binary for me"
- "Create a workflow to solve this crypto challenge"
- "Scan this web application for vulnerabilities"
- "Generate an exploit for this PWN challenge"