code-search

Search codebases via CodeGraph queries for symbols, callers, and dependencies.

Updated Apr 27, 2026
One-click install
npx skills add https://github.com/vTRKA/supervibe --skill code-search-vtrka
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-search
Source: https://github.com/vTRKA/supervibe/tree/main/skills/code-search
Command: npx skills add https://github.com/vTRKA/supervibe --skill code-search-vtrka

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires <resolved-supervibe-plugin-root>/scripts/supervibe-status.mjs, <resolved-supervibe-plugin-root>/scripts/search-code.mjs.

What problem does it solve?

It prevents confusing, risky coding sessions by helping you locate the relevant existing implementation, patterns, and callers in the codebase before making non-trivial changes.

Core Features & Use Cases

  • CodeGraph-powered discovery: Uses CodeGraph queries (symbols, edges, neighborhoods, and impact radius) to find where and how things are implemented and depended on.
  • Evidence-backed context: Produces agent-ready context packets that cite concrete file:line references and include graph evidence when structural claims are made.
  • Safety-first workflow discipline: Blocks or defers when the CodeGraph index is stale/partial or required proofs/receipts are missing, avoiding “stale hit as fact” errors.
  • Typical scenarios: Understanding “how X works,” identifying “where X is handled,” finding “who calls X,” and locating “similar patterns to Y” before refactoring or implementing changes.

Quick Start

Ask your AI to search the codebase for the callers and related context of a symbol named exactly where it’s defined, then return the top cited file:line results for your next step.

Frequently Asked Questions about code-search

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

FAQPage Schema
How do I find callers and locate implementation details before refactoring code?

To find callers and implementation details before refactoring, use CodeGraph search to query symbols, edges, and neighborhoods. This assesses the dependency blast radius and locates call sites, returning cited file:line references to prevent risky edits.

What is the best way to assess the impact radius of non-trivial code changes?

Assessing the impact radius of non-trivial code changes requires querying CodeGraph for structural edges and callers. This approach produces evidence-backed context packets with concrete file:line citations to verify dependency blast radius before editing.

How do I search a codebase to find similar patterns to an existing function?

Search a codebase for similar patterns by running a CodeGraph query to discover structural neighborhoods and top symbols. This source discovery method identifies matching implementation patterns and returns cited file:line evidence for verification.

Do I need a fresh CodeGraph index to search for code dependencies and callers?

Yes, you need a fresh CodeGraph index to search for code dependencies and callers. The workflow requires verifying index freshness first and will block or defer the search if the index is stale, partial, or missing required evidence.

Why does my code search return blocked results instead of caller information?

Your code search returns blocked results instead of caller information because the CodeGraph index is stale or partial. The safety-first workflow blocks execution when required proofs, receipts, or graph evidence are missing to avoid stale hit errors.