EclipseLlama 🦙

Topics: eclipse-plugin, ollama, local-llm, ai-coding-assistant, java, eclipse-ide

A privacy-first AI coding assistant for Eclipse IDE, powered by local Ollama or OpenAI models like CodeLlama and DeepSeek Coder. Features: AI chat, code explanation, bug fixes, test generation, Javadoc. 100% local, no cloud.

---

✨ Features

Feature	Description
💬 AI Chat	Conversation with your local LLM directly in Eclipse
🔍 Explain Code	Select code → Right-click → Get AI explanation
🐛 Fix Code	AI suggests fixes for selected code
🧪 Generate Tests	Auto-generate unit tests for your code
📝 Add Documentation	Generate Javadoc comments automatically
⚡ Smart Autocomplete	AI-powered code completion suggestions
🔒 100% Private	Everything runs locally, no cloud required

---

🚀 Quick Start

Prerequisites

1. OS: Windows 10/11, macOS, or Linux
2. Eclipse IDE 2023-12 or newer
3. Java 21 or compatible JDK
4. Ollama installed and running (Download) or Open AI compatible endpoint (like llama-server)

Install a Code Model

# Recommended models for coding
ollama pull codellama
# or
ollama pull deepseek-coder
# or
ollama pull qwen2.5-coder

Install the Plugin

1. Download EclipseLlama.jar from releases
2. Copy to Eclipse's dropins/ folder
3. Restart Eclipse
4. Complete the setup wizard

Install from Update Site

1. In Eclipse, go to Help → Install New Software...
2. Click Add...
3. Enter Name: EclipseLlama
4. Enter Location: https://vijaytank.github.io/eclipsellama/updatesite/
5. Click Add, select EclipseLlama, and follow the wizard.

---

⌨️ Keyboard Shortcuts

Shortcut	Action
Ctrl+Shift+L	Open EclipseLlama Chat
Ctrl+Shift+E	Explain Selected Code

---

🎯 Usage

Chat with AI

1. Press Ctrl+Shift+L or go to Help → EclipseLlama Chat
2. Type your question
3. Press Enter to send

Code Actions

1. Select code in the editor
2. Right-click → EclipseLlama submenu
3. Choose an action:
   • 🔍 Explain This Code
   • 🐛 Fix This Code
   • 🧪 Generate Tests
   • 📝 Add Documentation

Settings

Window → Preferences → EclipseLlama

• Configure Ollama endpoint URL
• Select default model
• Customize commit message prompts

---

🛠️ Configuration

Default settings are stored in ~/.eclipsellama/config.properties:

ollama.endpoint=http://localhost:11434
ollama.model=codellama

---

📋 Recommended Models

Model	Size	Best For
codellama	7B	General coding
codellama:13b	13B	Better accuracy
deepseek-coder	6.7B	Fast completions
qwen2.5-coder	7B	Multi-language

---

🛠️ Troubleshooting

Connection Timeouts

If you are using a large model on slower hardware, you might encounter connection timeouts. EclipseLlama now has an increased 60-second timeout to accommodate slower local LLM responses. If issues persist, ensure your Ollama or OpenAI endpoint is reachable.

---

🔒 Privacy

EclipseLlama runs entirely on your machine:

• ✅ No API keys required
• ✅ No data sent to cloud
• ✅ Works offline
• ✅ Your code stays private

🤝 Contributing

Issues and pull requests welcome!

---

📄 License

MIT License