Can it help with onboarding to a new repository?

Yes, it is designed to follow traces from entry points like package.json or main config files to help you understand how the system is wired together.

What happens if a feature is not found?

The skill provides a definitive 'not found' report, documenting the search strategies used (glob, grep, etc.) so you can confidently proceed knowing the feature does not exist.

How does this skill prevent AI hallucinations?

It enforces a 'verify, don't assume' workflow that requires Claude to use search and read tools to confirm the existence of files and code patterns before referencing them in a plan.

When should I use this instead of standard file reading?

Use this skill when you need to understand broad patterns, find where a feature is implemented, or verify if a specific design assumption matches the actual codebase state.

Codebase Investigator

Name: Codebase Investigator
Author: ed3dai

byed3dai

•

生产力与工作流

Conducts deep-dive analysis of existing codebases to verify architectural assumptions and establish factual patterns for implementation planning.

关于

The Investigating a Codebase skill empowers Claude to move beyond guesswork by grounding every design and implementation decision in the actual state of the repository. By systematically exploring entry points, following import traces, and verifying file locations through glob and grep strategies, it prevents hallucinations and ensures that new code integrates seamlessly with existing patterns. It is especially valuable during the planning phase of a feature to confirm that the assumed dependencies and structures actually exist before a single line of code is written.

主要功能

Design assumption verification to confirm if expected files and functions truly exist
Evidence-based reporting with exact file paths and line numbers
45 GitHub stars
Multi-strategy search using Glob patterns and Grep keywords to locate specific code patterns
Architectural trace mapping to follow imports and component relationships
Definitive negative verification to prevent hallucinating non-existent file paths

使用场景

Locating entry points and dependencies in unfamiliar repositories to map out component relationships
Finding and extracting existing code patterns to maintain architectural consistency in new modules
Verifying codebase structure before implementing a feature to ensure design assumptions are correct

关于

主要功能

Design assumption verification to confirm if expected files and functions truly exist
Evidence-based reporting with exact file paths and line numbers
45 GitHub stars
Multi-strategy search using Glob patterns and Grep keywords to locate specific code patterns
Architectural trace mapping to follow imports and component relationships
Definitive negative verification to prevent hallucinating non-existent file paths

使用场景

Locating entry points and dependencies in unfamiliar repositories to map out component relationships
Finding and extracting existing code patterns to maintain architectural consistency in new modules
Verifying codebase structure before implementing a feature to ensure design assumptions are correct