Ollama Copilot
🤖 An intelligent AI coding assistant powered by local Ollama models
Ollama Copilot is a VS Code extension that brings GitHub Copilot-like AI assistance directly to your editor, running entirely on your local machine using Ollama. Unlike simple chat extensions, this is a full agentic system with multi-mode support for planning, coding, and Q&A.
✨ Features
🎯 Three Intelligent Modes
- 💻 Code Mode: Full implementation workflow - plan → generate → preview → apply
- 📋 Plan Mode: Create detailed implementation plans without code generation
- 💬 Ask Mode: Context-aware Q&A about your codebase
🚀 Core Capabilities
- 🧠 Multi-Agent System: Specialized agents for planning, coding, and testing
- 🔍 Workspace Intelligence: Deep understanding of your project structure and tech stack
- 📝 Multi-File Editing: Make coordinated changes across multiple files
- 👀 Safe Previews: Review unified diffs before applying any changes
- ✅ Test Integration: Automatically run tests and iterate on failures
- 🎨 Modern UI: Clean sidebar chat interface similar to GitHub Copilot
- 📚 Chat History: Persistent conversation history across sessions
- ⏹️ Cancellation Support: Stop long-running operations at any time
- 🏠 100% Local: All AI processing happens on your machine via Ollama
- 🔒 Secure: Whitelisted command execution for safety
📋 Prerequisites
1. Install Ollama
Download and install Ollama from ollama.ai
2. Pull a Model
# Recommended: Qwen 2.5 Coder (7B)
ollama pull qwen2.5-coder:7b
# Or try other models:
ollama pull codellama:7b
ollama pull deepseek-coder:6.7b
ollama pull mistral:7b
3. Start Ollama Server
ollama serve
🔧 Installation
From VSIX (Recommended)
- Download the latest
.vsix file from Releases
- Open VS Code
- Press
Cmd+Shift+X (macOS) or Ctrl+Shift+X (Windows/Linux) to open Extensions
- Click the
... menu → Install from VSIX...
- Select the downloaded
.vsix file
- Reload VS Code when prompted
From Source
# Clone the repository
git clone https://github.com/YOUR_USERNAME/ollama-copilot.git
cd ollama-copilot
# Install dependencies
npm install
# Compile TypeScript
npm run compile
# Package extension
npx vsce package --allow-missing-repository
# Install the .vsix file in VS Code
🎯 Quick Start
Click the 🤖 Ollama Copilot icon in the Activity Bar (left sidebar)
2. Select Your Model
Use the dropdown at the top of the chat to choose your preferred Ollama model
3. Choose a Mode
- 💻 Code - Full implementation workflow
- 📋 Plan - Planning only
- 💬 Ask - Q&A with workspace context
4. Start Chatting!
Example prompts:
Code Mode:
"Add a login form with email validation"
"Refactor the UserService to use dependency injection"
"Create a REST API endpoint for user registration"
Plan Mode:
"How should I implement user authentication?"
"What's the best way to add caching to this API?"
Ask Mode:
"What does the fetchData function in api.ts do?"
"How is error handling implemented in this codebase?"
"Where is the database connection configured?"
🏗️ Architecture
ollama-copilot/
├── src/
│ ├── extension.ts # Main extension controller
│ ├── agents/ # AI agents
│ │ ├── planner.ts # Implementation planning
│ │ ├── coder.ts # Code generation
│ │ └── tester.ts # Test execution & fixing
│ ├── ollama/ # Ollama integration
│ │ ├── client.ts # HTTP client for Ollama API
│ │ └── modelSelector.ts # Model selection UI
│ ├── tools/ # Workspace operations
│ │ ├── workspace.ts # File I/O and structure
│ │ ├── search.ts # Code search
│ │ ├── patch.ts # Unified diff application
│ │ └── terminal.ts # Command execution
│ ├── ui/ # User interface
│ │ └── chatView.ts # Sidebar chat panel
│ └── protocol/ # Type definitions
│ ├── types.ts # TypeScript interfaces
│ └── prompts.ts # System prompts
└── docs/ # Documentation
⚙️ Configuration
Open VS Code Settings (Cmd+, or Ctrl+,) and search for "Ollama Copilot":
| Setting |
Default |
Description |
ollamaCopilot.apiUrl |
http://localhost:11434 |
Ollama API base URL |
ollamaCopilot.model |
qwen2.5-coder:7b |
Default Ollama model |
ollamaCopilot.temperature |
0.1 |
Model temperature (0-2) |
ollamaCopilot.maxTokens |
4000 |
Maximum tokens to generate |
ollamaCopilot.allowedCommands |
["npm test", ...] |
Whitelisted terminal commands |
🎨 Features in Detail
Code Mode: Full Implementation
- Planning Phase: AI analyzes your request and workspace to create an implementation plan
- Code Generation: Generates code changes across multiple files
- Preview: Shows unified diffs of all proposed changes
- Apply: Safely applies patches to your workspace
- Testing: Runs tests and iterates if failures occur
Plan Mode: Architecture Planning
- Creates detailed implementation plans without generating code
- Perfect for understanding how to approach complex features
- Lists files to modify, steps to take, and considerations
- Switch to Code mode to implement the plan
Ask Mode: Intelligent Q&A
- Context-aware answers based on your actual codebase
- Searches and reads relevant files automatically
- References specific code and file locations
- No code modifications, just helpful information
Chat History
- Conversations persist across VS Code sessions
- Stored locally in VS Code's global state
- Automatically trimmed to last 100 messages
- Clear chat anytime with dedicated button
- Cancel long-running operations
- Appears during AI processing
- Safe cancellation at natural breakpoints
- Resume with new prompts
🛠️ Development
Setup
npm install
npm run compile
Run & Debug
Press F5 in VS Code to launch the extension in a new Extension Development Host window
Package
npm run package
# Creates ollama-copilot-X.X.X.vsix
Lint
npm run lint
📖 Documentation
🤝 Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature)
- Commit your changes (
git commit -m 'Add some amazing feature')
- Push to the branch (
git push origin feature/amazing-feature)
- Open a Pull Request
📝 License
This project is licensed under the MIT License - see the LICENSE file for details.
🙏 Acknowledgments
Made with ❤️ for the open source community
