VSCodeCompass

AI-Powered Code Search and Question Answering for Your Codebase

CodeCompass is a VS Code extension that helps developers navigate and understand codebases through AI-powered search and question answering. Particularly useful for new team members trying to understand unfamiliar code.

Features

• Semantic Code Search - Find relevant files based on natural language queries
• AI-Powered Q&A - Ask questions about your codebase and get contextual answers
• Interactive Chat - Maintain conversations with context from previous questions
• Multiple AI Providers - Support for OpenRouter, Google Gemini, and local Ollama models

Installation

Prerequisites

From VSIX (Recommended)

1. Download latest .vsix from Releases
2. In VS Code: Extensions (Ctrl+Shift+X)
3. Click "..." > Install from VSIX...
4. Select downloaded file

Build from Source

# Clone repository
git clone https://github.com/balakumardev/VSCodeCompass.git
cd VSCodeCompass

# Install dependencies
npm install

# Build extension
npm run package

Setup

1. Set Up Qdrant Vector Database

Option A: Docker (Recommended)

docker run -d --name qdrant 
  -p 6333:6333 
  -v qdrant-data:/qdrant/storage 
  qdrant/qdrant

Option B: Native Installation

Follow the Qdrant installation guide.

Verify at: http://localhost:6333/dashboard

2. Configure AI Provider

OpenRouter (Cloud)

1. Create account at OpenRouter
2. Generate API key
3. In VS Code Settings:
   • Set codecompass.aiProvider to OPENROUTER
   • Set codecompass.openRouterApiKey to your key

Google Gemini (Cloud)

1. Get API key from Google AI Studio
2. In VS Code Settings:
   • Set codecompass.aiProvider to GEMINI
   • Set codecompass.geminiApiKey to your key

Ollama (Local)

1. Install Ollama
2. Pull a model: ollama pull llama3
3. In VS Code Settings:
   • Set codecompass.aiProvider to OLLAMA

Usage

Index Your Codebase

1. Open Command Palette (Ctrl+Shift+P)
2. Run CodeCompass: Index Workspace
3. Wait for indexing to complete

Search Code

1. Open Command Palette
2. Run CodeCompass: Search
3. Enter natural language query
4. Browse results and open files

Ask Questions

1. Open CodeCompass panel from sidebar
2. Type your question in the chat
3. Get AI-powered answers with code context

Configuration

Tech Stack

• Language: TypeScript
• Platform: VS Code Extension
• Vector DB: Qdrant
• AI Providers: OpenRouter, Gemini, Ollama
• Embeddings: Local or API-based

Architecture

+------------------+     +------------------+
|   VS Code UI     |     |    AI Provider   |
|  (Chat, Search)  |     | (OpenRouter/     |
+--------+---------+     |  Gemini/Ollama)  |
         |               +--------+---------+
         v                        |
+------------------+              |
|  CodeCompass     |<-------------+
|    Extension     |
+--------+---------+
         |
         v
+------------------+
|  Qdrant Vector   |
|    Database      |
+------------------+

Contributing

Contributions welcome! Please open issues or submit pull requests.

License

This project is open source. See the repository for license details.