explore

Searches a codebase with semantic queries to locate implementations and trace data flow.

Updated Jul 31, 2026
One-click install
npx skills add https://github.com/AarnavBaddam/skills --skill explore-aarnavbaddam
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: explore
Source: https://github.com/AarnavBaddam/skills/tree/main/explore
Command: npx skills add https://github.com/AarnavBaddam/skills --skill explore-aarnavbaddam

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Answering questions about how a codebase works often requires slow manual grepping and reading many files, which wastes time and context tokens when keyword search misses the relevant code. ## Core Features & Use Cases - Semantic Codebase Search: Translates a user question into natural language queries and runs the codebase_search tool from multiple angles. - Flow Tracing: Identifies entry points, data flow, and ownership of a behavior, then summarizes findings with concrete file paths. - Use Case: A user asks "where is authentication handled?" The skill runs 2-3 varied codebase_search queries, reads the returned sections, and summarizes the router, handlers, config, and tests involved. ## Quick Start Ask how a specific feature works in this repository and have the codebase searched to trace its implementation.

Frequently Asked Questions about explore

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find where a feature is implemented in a large codebase?

Use semantic codebase search with a natural language description of the functionality rather than keyword grep. Running 2-3 varied queries from different angles surfaces relevant files, which you then read to trace the implementation.

How to trace data flow through a repository?

Formulate a query like "Find the entry points and data flow for X, including router/handlers, config, and tests" and run it against the repo root. Read the returned sections and summarize the flow with concrete file paths.

When should I use semantic code search instead of grep?

Use semantic search when keyword grep is likely to miss relevant files, such as when you do not know exact identifier names. It works best for questions about how something works or where behavior is owned.

Why does a single codebase search query miss relevant files?

One query captures only one phrasing of the concept, so relevant code described with different terminology is missed. Running 2-3 queries with different search strings explores the codebase from multiple angles and improves coverage.