Patchy FAQs

Question 1

Which ProjectDiscovery tools are included in Patchy?

Accepted Answer

Patchy wraps six core ProjectDiscovery tools: subfinder for passive subdomain enumeration, dnsx for DNS resolution, httpx for HTTP probing, naabu for TCP port scanning, katana for web crawling, and nuclei for template-based vulnerability scanning.

Question 2

Can Patchy be easily set up and updated?

Accepted Answer

Yes, Patchy supports self-bootstrapping, allowing it to install all necessary ProjectDiscovery tools and Nuclei templates automatically on its first run. It also provides easy configuration via environment variables or a YAML file.

Question 3

How does Patchy ensure secure and safe execution for LLM agents?

Accepted Answer

Patchy enforces a 'deny-by-default' policy, requiring explicit scope configuration. It utilizes sandboxed execution, a binary allowlist, and bounds output to prevent misuse, ensuring security tasks are performed safely within defined boundaries.

Question 4

What is Patchy?

Accepted Answer

Patchy is an MCP server that integrates core ProjectDiscovery security tools (like Nuclei, Subfinder, Katana) as typed, policy-enforced, and sandboxed services specifically designed for large language model (LLM) agents.

Question 5

What is the Model Context Protocol (MCP) and how does Patchy use it?

Accepted Answer

The Model Context Protocol (MCP) is a standard for LLM agents to interact with external tools. Patchy exposes all integrated ProjectDiscovery tools as structured MCP services, enabling seamless and secure integration with MCP-compatible LLM agents like Claude.

Patchy

Patchy

主要功能

使用案例

主要功能

使用案例