Gedit LLaMA Plugin

A Gedit plugin that integrates with openai API compatible local LLM servers (like llama.cpp) to ask questions about selected text.

Features

• Context-Aware Prompts: Automatically includes selected text in your prompt when asking LLaMA questions.
• Streaming Support: Displays responses as they arrive, providing real-time output from the model.
• Customizable Configuration: Easily configure API URL, API key, model name and keyboard shortcut.
• Multi-line Prompt Input: Use a multi-line text area to compose complex prompts.
• Copy To Clipboard: button to copy LLM response to clipboard.

Requirements

• Gedit 44+
• Python 3.x
• requests library (install with pip install requests or via your distro's package manager python3-requests)

Installation

1. Install the schema:

   cp org.gnome.gedit.plugins.gedit_llama.gschema.xml ~/.local/share/glib-2.0/schemas/
   glib-compile-schemas ~/.local/share/glib-2.0/schemas/

2. Copy plugin files:

   mkdir -p ~/.local/share/gedit/plugins
   unzip gedit_LLaMA.zip -d ~/.local/share/gedit/plugins/

3. Enable the plugin:

   • Open Gedit
   • Go to Edit → Preferences → Plugins
   • Enable "Gedit LLaMA"

Usage

1. Select text in your document (optional)
2. Right-click in the editor and choose Gedit LLaMA → Ask LLaMA
3. Enter a prompt in the dialog
4. View results in a popup dialog that shows real-time streaming output

Configuration

You can customize:

• API URL (default: http://127.0.0.1:5000/v1/chat/completions)
• API Key (if required by your server)
• Model name (default: llama.cpp)
• Keyboard shortcut (default: <Ctrl><Alt>l)

Access the configuration via:

• Right-click menu → Configure LLaMA

How It Works

1. Select text in your document
2. Right-click and select "Ask LLaMA"
3. Enter your question or instruction
4. Plugin sends selected text (if any) + prompt to your local LLM server
5. Response is displayed in a streaming popup dialog

Example Use Cases

• Explain selected code snippets
• Generate documentation for code
• Find bugs or suggest improvements
• Summarize selected text
• Translate code comments
• Debugging assistance
• Code generation based on context

Notes

• Requires a local LLM server like llama.cpp running at the configured URL
• Supports both streaming and non-streaming responses
• Plugin automatically detects when new tabs are opened and connects to their views
• Uses GSettings for persistent configuration storage

Troubleshooting

If you encounter issues:

1. Ensure your local LLM server is running and accessible at the specified URL
2. Verify that requests is installed (pip install requests)
3. Check that the schema file was properly compiled using glib-compile-schemas
4. Confirm the plugin is enabled in Gedit's preferences

License

MIT License - see LICENSE file for details.