An AutoHotkey v2 clipboard manager that transforms your clipboard history into context for LLM conversations.

LLMClip monitors your clipboard and aggregates copied text, files, and folders into a unified context for AI chat. It seamlessly integrates with various LLM providers (Groq, Ollama, Google Gemini, OpenAI) to allow you to chat with your clipboard content.

• Smart Capture: Automatically captures copied text, file paths, and folder contents.
• Context Awareness:
  • Files & Folders: Reads content of text-based files and lists folder contents.
  • PDF Support: Auto-extracts text and images from PDF files (requires pdftotext and pdfimages).
  • Images: Supports clipboard images and image files for multimodal LLMs.
  • URLs: Extracts content and metadata from web links.
• Selection Control: Checkboxes allow you to select specific items from your history to send as context.

• Multi-Provider: Pre-configured support for Groq, Ollama, Google (Gemini), and OpenAI.
• Customizable: Easy settings.json configuration for models, system prompts, and API keys.
• Tools: Capable of executing PowerShell commands for file system operations when enabled.

• Markdown Rendering: chat history is rendered with Markdown, including syntax highlighting for code blocks.
• History Management:
  • Compression: "Compress" feature to summarize long conversations and save tokens.
  • Extract Notes: "Extract Learnings" to summarize key information from a session.
  • Editing: Edit previous user messages and regenerate responses.
• Session Management: Switch between multiple independent chat sessions.

1. Install AutoHotkey v2: Download from autohotkey.com.
2. Install cURL: Ensure curl is installed (winget install cURL.cURL or via curl.se).
3. PDF Tools (Optional): For PDF support, place pdftotext.exe and pdfimages.exe in a pdf/ folder within the project directory. (These are part of Xpdf or Poppler utils).
4. Clone Repository: Clone or download this project.
5. Configuration: Create a settings.json file (see below).
6. Run: Execute clip.ahk.

Create a settings.json file in the project root. Here is a comprehensive example:

{
    "selectedLLMType": "groq",
    "providers": {
        "groq": {
            "curl": "curl -s -S -X POST \"https://api.groq.com/openai/v1/chat/completions\" -H \"Content-Type: application/json\" -H \"Authorization: Bearer <<KEY>>\" -d \"@{1}\" -o \"{2}\"",
            "model": "llama-3.3-70b-versatile",
            "temperature": 0.7,
            "system_prompt": "You are a helpful assistant. Be concise and direct in your responses."
        },
        "ollama": {
            "curl": "curl -s -S -X POST \"http://localhost:11434/api/chat\" -H \"Content-Type: application/json\" -H \"Authorization: Bearer <<KEY>>\" -d \"@{1}\" -o \"{2}\"",
            "model": "llama3",
            "temperature": 0.7,
            "system_prompt": "You are a helpful assistant. Be concise and direct in your responses."
        }
    }
}

Replace <<KEY>> with your actual API keys.

• F3: Toggle Recording / Show Window / Stop Recording cycle.
• Enter: Send message.
• Shift+Enter: Insert newline.

• Context List: Uncheck items you don't want to send to the LLM.
• Images: If an image is in the clipboard or selected context, it will be sent to multimodal models (like Gemini or GPT-4o).
• Tools: Enable the ps1 checkbox to allow the LLM to write and execute PowerShell scripts (Use with caution!).

MIT License

Uses libraries from thqby/ahk2_lib