🦙 Ollama Agent - Local AI Coding Assistant

Transform your locally-running Ollama models into an intelligent coding agent with capabilities similar to GitHub Copilot or Cursor—all running 100% locally with complete privacy. No cloud, no subscriptions, no data leaks.

✨ Features

🤖 Intelligent Agent Mode

• Autonomous Task Execution: Agent can read files, search codebases, create/edit files, and run tasks
• Multi-Step Reasoning: Up to 8-step problem-solving with context awareness
• Smart Tool Usage: Automatically explores your project structure and applies changes
• Read & Agent Modes: Choose between safe read-only or full agent capabilities

💬 Interactive Chat Interface

• Modern WebView UI: Clean, responsive chat interface with streaming responses
• Multi-Model Support: Chat with multiple Ollama models simultaneously
• Context-Aware: Automatically includes selected code, open files, and project structure
• Conversation History: Keep track of all interactions within sessions

📝 Code Generation & Editing

• Inline Code Generation: Generate code directly at cursor position
• Smart Refactoring: Simplify, optimize, or add types to selected code
• Function Extraction: Automatically extract functions with AI assistance
• Code Translation: Convert code between languages

📚 Documentation & Analysis

• README Generator: Dedicated mode to generate comprehensive project documentation
• Docstring Completion: AI-generated documentation for functions and classes
• Code Explanation: Understand complex code with AI explanations on hover
• PR Summaries: Generate pull request descriptions and review comments

🧪 Testing & Quality

• Unit Test Generation: Create comprehensive test suites automatically
• Problem Analysis: AI-powered debugging and error resolution
• Code Review: Get suggestions for improvements and best practices

🛠️ Developer Tools

• Project Indexer: Fast semantic search across your entire codebase
• Commit Message Generation: Smart git commit messages from staged changes
• Shell Command Explanation: Understand complex terminal commands
• File Creation with AI: Generate boilerplate files with context

🎨 Customization

• Configurable Settings: Host, port, system prompts, and behavior modes
• Keyboard Shortcuts: Quick access to all commands
• Context Menu Integration: Right-click access in editor
• Status Bar Integration: Quick model switching and status updates

📦 Installation

Prerequisites

1. Install Ollama: Download from ollama.ai
2. Pull Models: Choose your preferred models

   ollama pull llama3.1:8b
   ollama pull mistral
   ollama pull codellama
   ollama pull deepseek-coder

3. Start Ollama: Ensure it's running on localhost:11434

   ollama serve

Install Extension

From VS Code Marketplace (Coming Soon):

1. Open VS Code
2. Go to Extensions (Ctrl+Shift+X / Cmd+Shift+X)
3. Search for "Ollama Agent"
4. Click Install

From VSIX (Manual):

1. Download the .vsix file from GitHub Releases
2. Open VS Code
3. Run command: Extensions: Install from VSIX...
4. Select the downloaded file

🚀 Quick Start

1. Open Chat Panel

• Click the Ollama icon in the Activity Bar (left sidebar)
• Or run command: Ollama Agent: Open Chat
• Or use keyboard shortcut: Ctrl+Shift+P → "Ollama Agent"

2. Select Agent Mode

• Read Mode (Safe): AI can only read and suggest changes
• Agent Mode (Powerful): AI can automatically create/edit files

3. Start Chatting

• Type your question or request
• Select one or more models to query
• Get streaming responses in real-time

4. Use Context Menu Commands

• Select code → Right-click → Ollama Agent options
• Ask about selection, generate code, refactor, explain, and more

⚙️ Configuration

Settings

Access via File > Preferences > Settings → Search "Ollama Agent"

Setting	Default	Description
ollamaAgent.host	localhost	Ollama server host
ollamaAgent.port	11434	Ollama server port
ollamaAgent.systemPrompt	You are a helpful coding assistant.	Default system prompt
ollamaAgent.provider	ollama	AI provider (ollama/openai/anthropic)
ollamaAgent.mode	read	Interaction mode (read/agent)
ollamaAgent.enableInlineCompletions	false	Experimental: inline completions
ollamaAgent.enableExplainOnHover	false	Experimental: hover explanations
ollamaAgent.enableDocstringCompletion	false	Experimental: docstring suggestions

Example Configuration

{
  "ollamaAgent.host": "localhost",
  "ollamaAgent.port": 11434,
  "ollamaAgent.mode": "agent",
  "ollamaAgent.systemPrompt": "You are an expert TypeScript developer focused on clean code and best practices."
}

📖 Usage Examples

Generate a Function

1. Place cursor where you want the function
2. Run: Ollama Agent: Generate Code Here
3. Describe what you want: "Create a function to validate email addresses"
4. Review and insert the generated code

Refactor Code

1. Select the code to refactor
2. Right-click → Ollama Agent: Brush – Simplify
3. AI will suggest cleaner, more readable code
4. Review changes in diff view and apply

Create Unit Tests

1. Select a function or class
2. Run: Ollama Agent: Generate Unit Tests
3. AI generates comprehensive test suite
4. Tests appear in new file or clipboard

Generate README

1. Run: Ollama Agent: README Generator
2. Choose style (Technical/Product/Library)
3. Choose placement (GitHub/Marketplace/Website)
4. Add custom notes
5. Enable deep scan for comprehensive analysis
6. AI generates complete README with all sections

Debug with AI

1. Click on a problem in Problems panel
2. Run: Ollama Agent: Fix This Problem
3. AI analyzes error and suggests fixes
4. Apply fix with one click

🎯 Command Reference

Chat & Interaction

• Ollama Agent: Open Chat - Open main chat panel
• Ollama Agent: README Generator - Generate project documentation
• Ollama Agent: Ask About Selection - Query AI about selected code

Code Generation

• Ollama Agent: Generate Code Here - Insert AI-generated code
• Ollama Agent: Generate Unit Tests - Create test suite
• Ollama Agent: Generate Commit Message - Smart git messages
• Ollama Agent: Generate README - Auto-generate documentation

Code Editing

• Ollama Agent: Edit Selection by Instruction - Guided refactoring
• Ollama Agent: Brush – Simplify - Simplify code
• Ollama Agent: Brush – Add Types - Add TypeScript types
• Ollama Agent: Brush – Optimize - Performance optimization
• Ollama Agent: Extract Function - Extract selected code

Analysis & Explanation

• Ollama Agent: Explain Selection - Understand code
• Ollama Agent: Analyze Problems - Debug assistance
• Ollama Agent: Explain Command - Shell command help
• Ollama Agent: Show Index Stats - View codebase stats

Project Management

• Ollama Agent: PR Summary - Generate PR descriptions
• Ollama Agent: PR Review Comments - AI code review
• Ollama Agent: Rebuild Project Index - Refresh semantic search
• Ollama Agent: Add File (AI) - Generate boilerplate files

Translation

• Ollama Agent: Translate Selection - Convert between languages

🔧 Troubleshooting

Ollama Not Connecting

Problem: "Failed to connect to Ollama"

Solutions:

1. Verify Ollama is running: ollama serve
2. Check settings: host=localhost, port=11434
3. Test connection: curl http://localhost:11434/api/tags
4. Restart VS Code

No Models Available

Problem: "No models found"

Solutions:

1. Pull a model: ollama pull llama3.1
2. List models: ollama list
3. Refresh extension: Run Developer: Reload Window

Slow Response Times

Problem: AI responses are very slow

Solutions:

1. Use smaller models (e.g., llama3.1:8b instead of 70b)
2. Reduce context: Disable deep scan in README generator
3. Close other resource-intensive applications
4. Consider using GPU acceleration with Ollama

Agent Mode Not Working

Problem: Agent mode doesn't apply changes

Solutions:

1. Check mode setting: ollamaAgent.mode = agent
2. Verify file permissions (read/write access)
3. Check workspace trust: File → Trust Workspace
4. Review error messages in Output panel

Extension Not Loading

Problem: Extension doesn't activate

Solutions:

1. Check VS Code version (minimum 1.105.0)
2. View logs: Output → Ollama Agent
3. Disable conflicting extensions
4. Reinstall extension

🛠️ Development

Setup

# Clone repository
git clone https://github.com/IamNishant51/Ollama-Agent-vs-code-extension.git
cd Ollama-Agent-vs-code-extension

# Install dependencies
npm install

# Build webview
npm run build:webview

# Compile TypeScript
npm run compile

# Watch mode (auto-recompile)
npm run watch

Publishing

See PUBLISHING.md for a step-by-step guide to package and publish this extension to the VS Code Marketplace.

Testing

# Run linter
npm run lint

# Run tests
npm test

Debug

1. Open project in VS Code
2. Press F5 to launch Extension Development Host
3. Test features in new window
4. View logs in Debug Console

🤝 Contributing

Contributions are welcome! Please read our Contributing Guide for details.

Areas for Contribution

• Add support for more AI providers
• Improve agent reasoning capabilities
• Add more code generation templates
• Enhance UI/UX with additional themes
• Write comprehensive tests
• Improve documentation
• Report bugs and suggest features

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Ollama for providing the local LLM runtime
• VS Code for the excellent extension API
• The open-source community for inspiration and support

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: Create an issue for support

🗺️ Roadmap

• Multi-file editing in single operation
• Voice input support
• Custom prompt templates
• Extension marketplace for agent tools
• Collaborative coding sessions
• Code review workflows
• Integration with popular dev tools
• Mobile companion app

⭐ Star History

If you find this extension useful, please consider giving it a star on GitHub!

---

Made with ❤️ by Nishant Unavane

Powered by Ollama • Privacy-First • Open Source