What is 'orthogonal exploration' in this context?

It is a strategy that searches for code from multiple distinct angles—such as how a feature is implemented versus how it is used—to ensure comprehensive coverage without overlap.

How does the skill handle duplicate findings from different agents?

It includes a mandatory cross-reference phase that identifies duplicates and merges overlapping line ranges into a single, clean entry in the final reading list.

When should I use the 'Very-Thorough' level?

Use this level for broad queries like 'architecture overview' or 'security audit' where missing a single configuration file or edge-case test could lead to incomplete understanding.

Does this skill read every file in my repository?

No, it orchestrates specialized agents to identify the most relevant files and line ranges. The final synthesis provides a list for the main agent or user to read, optimizing context window usage.

Codebase Explorer & Navigator

Name: Codebase Explorer & Navigator
Author: doodledood

bydoodledood

•

開発者ツール

Orchestrates parallel sub-agents to systematically map complex codebases through implementation, usage, and test-based exploration.

概要

The Explore Codebase skill automates the discovery phase of development by launching multiple specialized agents to analyze a project from orthogonal perspectives. Whether you are performing an architectural audit or onboarding to a new repository, this skill coordinates agents to find core logic, call sites, test coverage, and configuration files relevant to your query. It synthesizes these findings into a unified, prioritized reading list with specific line ranges, ensuring that cross-file interactions and hidden dependencies are never overlooked during the development process.

主な機能

Parallel agent orchestration for thorough codebase coverage
8 GitHub stars
Unified synthesis of agent findings into prioritized reading lists
Auto-selection of thoroughness levels based on query complexity
Orthogonal exploration strategies (Implementation, Usage, Tests, Config)
Automated gap analysis and cross-reference deduplication

ユースケース

Identifying all relevant files and test cases for an architectural audit
Mapping out a complex feature's architecture and call hierarchy
Onboarding to unfamiliar repositories by discovering core entry points and data flows

概要

主な機能

Parallel agent orchestration for thorough codebase coverage
8 GitHub stars
Unified synthesis of agent findings into prioritized reading lists
Auto-selection of thoroughness levels based on query complexity
Orthogonal exploration strategies (Implementation, Usage, Tests, Config)
Automated gap analysis and cross-reference deduplication

ユースケース

Identifying all relevant files and test cases for an architectural audit
Mapping out a complex feature's architecture and call hierarchy
Onboarding to unfamiliar repositories by discovering core entry points and data flows