Code Search | Grep by Vercel
Effortlessly search for code, files, and paths across a million GitHub repositories.
Developer Tool Announcement
Summary
The blog post introduces Vercel’s Model Context Protocol (MCP) server that lets AI agents query over a million public GitHub repositories. It explains how the HTTP‑based API can retrieve pattern‑matched code snippets and offers fast, relevant results for implementation or error‑handling queries.
Highlights
- Enables AI agents to search more than one million GitHub repos via Grep’s MCP server.
- Supports pattern‑based queries, language filters, and relevance ranking.
- Provides a simple HTTP API that integrates with agents like Cursor and Claude.
- Designed for rapid code discovery, debugging assistance, and learning from open‑source patterns.
auto-generated
Context
Audience
Developers building AI‑assisted coding tools, platform engineers interested in large‑scale code search, and technical leads evaluating AI integration for code bases
DomainSoftware Development / AI‑augmented coding
Formatblog article
Accessfree online
Source Overview
Grep's MCP server enables AI applications to search and retrieve code snippets from over a million public GitHub repositories with pattern-based queries and language filters.
The server provides fast, HTTP-based access using the Model Context Protocol, allowing integration with agents like Cursor and Claude. Results are ranked for relevance and can be used to solve implementation or error-handling questions programmatically.
Topics
Related
Vercel Model Context Protocol documentationCursor AI code editorClaude by AnthropicGitHub Code SearchOpenAI Codex
