How does the 'test agent' help in exploration?

The test agent analyzes test structures and fixtures to reveal the actual runtime behavior and edge cases, which are often more descriptive than source code alone.

When should I use dev-explore instead of dev-plan?

Use dev-explore first to understand how the current system works; use dev-plan afterward to define the specific implementation steps for a change.

Does this skill work on specific folders or entire repositories?

It can do both. You can trigger a global exploration for the whole project or use Focus Mode to analyze a specific feature or directory.

What is the main benefit of using Codebase Explorer?

It prevents 'tunnel vision' by using parallel agents to analyze structure, data, behavior, and tests simultaneously, providing a holistic view faster than manual reading.

Codebase Explorer

Name: Codebase Explorer
Author: phbpx

byphbpx

0•

Learning & Documentation

Systematically explores and maps unfamiliar codebases using parallel AI agents to build comprehensive architectural mental models.

Codebase Explorer is a specialized Claude Code skill designed to eliminate the initial friction of working with new or complex projects. By dispatching multiple specialized agents to analyze directory structures, data flows, core behaviors, and testing patterns simultaneously, it provides a high-level architectural summary without the "tunnel vision" of manual file reading. It is an essential tool for developers onboarding to new teams, auditing legacy code, or preparing for significant refactoring tasks by identifying key files, design patterns, and potential pitfalls before a single line of code is changed.

Key Features

01Automated Data Flow and System Dependency Mapping

02Structured Synthesis of Architecture and Module Organization

03Identification of Non-Obvious Design Decisions and 'Footguns'

04Targeted Focus Mode for specific feature area deep-dives

05Parallel Multi-Agent Discovery (Structure, Data, Behavior, Tests)

060 GitHub stars

Use Cases

01Auditing legacy codebases to understand hidden dependencies and logic

02Pre-planning architectural changes by mapping current system behavior

03Rapid onboarding when joining a new project or repository

Key Features

01Automated Data Flow and System Dependency Mapping

02Structured Synthesis of Architecture and Module Organization

03Identification of Non-Obvious Design Decisions and 'Footguns'

04Targeted Focus Mode for specific feature area deep-dives

05Parallel Multi-Agent Discovery (Structure, Data, Behavior, Tests)

060 GitHub stars

Use Cases

01Auditing legacy codebases to understand hidden dependencies and logic

02Pre-planning architectural changes by mapping current system behavior

03Rapid onboarding when joining a new project or repository