grepai

Perform semantic code search and trace call relationships across codebases.

4|Updated Jan 21, 2026
One-click install
npx skills add https://github.com/audibleblink/skills --skill grepai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: grepai
Source: https://github.com/audibleblink/skills/tree/main/grepai
Command: npx skills add https://github.com/audibleblink/skills --skill grepai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Semantic code search and call graph tracing help developers find relevant code by meaning, map relationships between functions, and understand codebases faster than traditional keyword searches.

Core Features & Use Cases

  • Semantic search to locate code by intent rather than exact text.
  • Call graph tracing to identify callers and callees and map dependencies.
  • Output-friendly formats (JSON/TOON) for integration with AI agents and tooling.

Quick Start

Initialize the project, build the index, and run a sample search to explore results.

Frequently Asked Questions about grepai

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

FAQPage Schema
How does semantic code search differ from traditional keyword matching?

Semantic code search locates relevant code by intent and meaning rather than exact text, enabling developers to find functions based on what they do instead of specific keywords or variable names.

How do I trace callers and callees to map dependencies across a large repository?

Call graph tracing identifies callers and callees by extracting function relationships across a codebase, allowing you to map dependencies and understand execution flow for debugging and onboarding.

Do I need a running embedding provider to perform semantic search on my codebase?

Yes, semantic search requires a running embedding provider to build a local code index. You must initialize the project, build the index, and run a sample search to explore results.

Can I output call graph extraction results in JSON or TOON format for AI agent integration?

Yes, the tool supports JSON and TOON output formats for call graph extraction and semantic search results, enabling integration with AI agents and downstream tooling pipelines.

What is the best way to understand a large unfamiliar codebase for onboarding?

Combining semantic search with call graph tracing maps function relationships and locates code by intent, accelerating codebase understanding and onboarding faster than traditional keyword searches.

What are the limitations of semantic code search when applied to large repositories?

Semantic search on large repositories requires maintaining a local code index and a running embedding provider, which adds environment setup overhead and depends on index freshness for accurate results.