cursor-codebase-indexing_skill

Overview

This skill sets up and optimizes Cursor codebase indexing to enable fast semantic search and improved AI context for your project. It walks through enabling indexing, creating exclusion rules, monitoring progress, and validating results so your Cursor workspace becomes queryable with @codebase. Use this skill to turn a code repository into a searchable, AI-aware knowledge source.

How this skill works

The skill inspects your project workspace and builds a searchable index of source files, symbols, and definitions. It guides you to enable indexing in Cursor, add a .cursorignore file to exclude large or irrelevant paths, and monitor the status bar until indexing finishes. Once complete, the index powers @codebase queries and semantic code search inside Cursor.

When to use it

• Enabling semantic code search for a new Cursor project
• Improving AI suggestions by giving the agent structured project context
• Reducing irrelevant results by excluding large generated folders
• After cloning or pulling large repositories that need fresh indexing
• Before running cross-file code analysis or automated refactoring tasks

Best practices

• Create a .cursorignore at the project root and exclude node_modules, build, dist, and other generated folders
• Prioritize indexing only source directories to save disk and speed up queries
• Ensure Cursor is authenticated and you have enough disk space before starting indexing
• Monitor the status bar and retry if network issues interrupt initial indexing
• Test with simple @codebase queries after indexing to validate results

Example use cases

• Search for function definitions and usages across a monorepo using @codebase queries
• Improve code review context by letting AI reference exact symbol definitions
• Exclude large test fixtures so semantic search returns relevant production code
• Rebuild the index after major refactors to keep AI suggestions accurate
• Set up indexing in CI environments to support automated analysis tools