Index your project to enable powerful AI-powered semantic code search.
What is Codebase Indexing?
Codebase indexing creates an AI-powered searchable index of your project files:
- Semantic search: Find code by describing what it does
- Concept matching: Search for code patterns by concept
- AI-assisted discovery: The Embedr Agent can automatically search your codebase
Enable Auto-Indexing
- Open your project in the Editor
- Navigate to Settings tab (gear icon in toolbar)
- Scroll to Codebase Indexing section
- Toggle Enable Auto-Indexing
When enabled, the codebase is automatically indexed when you open a project and stays in sync with file changes.
Manual Indexing
- Go to Settings → Codebase Indexing
- Click Index Project (or Re-index Project)
- Wait for indexing to complete
To remove indexed data, click Clear Index.
Using Semantic Search
Search Palette (Keyboard Shortcuts)
| Platform | Shortcut |
|---|---|
| macOS | ⌘+Shift+F or ⌘+P |
| Windows | Ctrl+Shift+F or Ctrl+P |
Search Modes
| Mode | Description |
|---|---|
| Hybrid | Combines semantic and text search for best accuracy |
| Semantic | AI-powered conceptual search |
| Quick | Fast text-only search |
AI Agent Search
Ask the Embedr Agent to search:
- "Find the code that handles WiFi connection"
- "Where is the button debounce logic?"
- "Show me functions related to sensor calibration"
Index Status
View your index status in Settings → Codebase Indexing:
- Files Indexed: Number of files in the index
- Total Chunks: Number of searchable code chunks
Supported File Types
- Arduino sketches (
.ino) - C/C++ files (
.c,.cpp,.h,.hpp) - Markdown documentation (
.md) - JSON configuration files (
.json)
Tips
- Index your project before asking the AI to help with code tasks
- Use specific, descriptive queries for best results
- Re-index after major refactoring or adding many new files