Ferret FAQs

Question 1

What kind of reports can Ferret generate?

Accepted Answer

Ferret offers both static analysis reports (repository overview, dependency graphs, architectural breakdown) and AI-powered 'deep' reports providing a 10-section expert analysis covering architecture, data flow, strengths, risks, and learning takeaways from your codebase.

Question 2

Does Ferret require an AI or LLM to function?

Accepted Answer

No, Ferret's static analysis tools (like 'scan', 'dependencies', 'architecture') are free and do not require an LLM. Its AI-powered tools ('deep' and 'ask') do require an LLM, which can be configured with Anthropic, OpenAI, or local LLMs like Ollama.

Question 3

What is Ferret and what does it do?

Accepted Answer

Ferret is an MCP (Model Context Protocol) server that extracts complete knowledge from any codebase. It combines static analysis to detail architecture, patterns, dependencies, and API surface, with AI-powered tools for deep interpretation and contextual Q&A.

Question 4

How does Ferret integrate with existing developer tools?

Accepted Answer

As an MCP server, Ferret seamlessly integrates with various MCP clients such as Claude Code, Claude Desktop, and Cursor. You can also configure it to use local LLMs, providing flexibility for your specific development environment.

Question 5

What is the 'ask' tool?

Accepted Answer

The 'ask' tool is an AI-powered feature that allows you to ask any question about a repository. Ferret will provide an answer using its full codebase context, similar to having an expert senior engineer explain your project.

Ferret

Ferret

主要功能

使用案例

主要功能

使用案例