Codebase Excavator FAQs

Question 1

Does this skill work with any programming language?

Accepted Answer

Yes, because it analyzes entry points and file relationships through code traversal and grep patterns, it is language-agnostic and works with most web, CLI, and backend frameworks.

Question 2

Can I document a specific part of the system instead of the whole thing?

Accepted Answer

Yes, you can target specific areas using command arguments like feature=name or entry=/api/path to focus the excavation on a single module.

Question 3

What is the 'Verification Checklist' step?

Accepted Answer

It is a final audit that compares documented features against actual code routes and files to ensure no hidden API endpoints or backend-only features were missed during discovery.

Question 4

Where is the generated documentation stored?

Accepted Answer

All generated feature specifications are stored as markdown files in the .lore/reference/ directory, with a centralized index maintained in .lore/excavations/index.md.

Question 5

How does the Excavate skill discover features?

Accepted Answer

It uses a surface-surveyor agent to identify entry points like API routes and UI components, then traces the code logic to identify related files, data models, and dependencies.

Codebase Excavator

Key Features

Use Cases

Codebase Excavator

Key Features

Use Cases