PatchMind IDE

PatchMind is an AI-enhanced code editor with inline LLM support, real-time prompt workflows, and RAG-based file context awareness.

✨ Features

• LLM Chat Integration — Chat with Gemini or Ollama-backed models
• Inline Editing — Apply prompts directly to open documents
• Prompt Manager — Save, reuse, and batch-apply prompts to files
• Real-Time Context — File tree with checkboxes for RAG context control
• Token Awareness — Visual context token tracking and budgeting
• Themes & Fonts — Customizable editor appearance
• Multi-Provider Support — Easily switch between LLMs

🔧 Requirements

• Python 3.11+
• Poetry (Recommended for development) or pip
• An Ollama instance running (locally or via Docker)

🚀 Running Locally

There are two main ways to install and run PatchMind locally:

1. Using Poetry (Recommended)

# Clone the repository (if you haven't already)
# git clone <repository_url>
# cd patchmind-ide

# Install dependencies using Poetry
poetry install --no-dev

# Run the application
poetry run python -m pm

2. Using Pip

# Clone the repository (if you haven't already)
# git clone <repository_url>
# cd patchmind-ide

# Install dependencies using pip
pip install -r requirements.txt

# Run the application
python -m pm

🐳 Running with Docker Compose

This method runs both the PatchMind IDE and an Ollama instance in separate containers using Docker Compose. This is useful for isolating dependencies and ensuring a consistent environment.

Prerequisites:

• Docker (Install Docker)
• Docker Compose (Install Docker Compose)
• An X Server running on your host machine (this is standard on most Linux desktops).

Steps:

1. Allow Container Access to X Server: You need to explicitly grant containers permission to connect to your host's X Server for the GUI to display. Open a terminal on your host machine and run:

   xhost +local:docker

   (Note: This command allows any local user running Docker containers to connect to your X server. This is generally safe for local development but be aware of the security implications if you run untrusted containers. You can revoke permission later with xhost -local:docker)

2. Build and Run: Navigate to the project's root directory (where docker-compose.yml is located) in your terminal and run:

   docker compose up --build

   This command will:

   • Build the patchmind-ide Docker image based on the Dockerfile (if not already built).
   • Download the ollama/ollama image (if not already present).
   • Start both the patchmind-ide and ollama containers.
   • The PatchMind IDE GUI should appear on your desktop.

3. Using Ollama:

   • The IDE container is configured to connect to Ollama at http://ollama:11434.
   • You will need to pull models into the Ollama container after it starts. You can do this by opening another terminal and running:

     docker exec -it patchmind-ollama ollama pull llama3:8b
     # Or any other model you need

   • The IDE should then be able to list and use the models pulled into the ollama container.

4. Stopping:

   • Press Ctrl+C in the terminal where docker compose up is running.
   • To ensure containers are fully stopped and removed, run:

     docker compose down

   • Ollama models will be persisted in the ./ollama_data volume on your host.

Troubleshooting Docker:

• GUI doesn't appear / X11 Connection Error: Ensure xhost +local:docker was run correctly and your DISPLAY environment variable is set properly on the host. Check the container logs (docker logs patchmind-ide-app). You might need to adjust volume mounts for X11 authentication (XAUTHORITY) depending on your system setup.
• Cannot connect to Ollama: Verify the ollama container is running (docker ps). Check the OLLAMA_HOST environment variable in the docker-compose.yml file and ensure the IDE's Ollama client uses it (this might require code adjustments if the client library doesn't automatically pick up OLLAMA_HOST).

🧠 Development

• Auto-format: black .
• Lint: ruff .

🗂 Project Structure

pm/
├── core/               # Logic, config, services
├── ui/                 # Qt widgets & dialogs
├── handlers/           # Connects UI events to Core logic
├── assets/             # QSS stylesheets, etc.
├── __main__.py         # Entry point
Dockerfile              # For building the Docker image
docker-compose.yml      # For running with Docker Compose
pyproject.toml          # Poetry config
README.md               # This file
...

📋 Guidelines

• Follows PEP8 with black & ruff
• Use Signal/Slot correctly with type annotations
• Never modify the GUI from non-main threads — use signals
• See DEVELOPER_GUIDELINES.md for more details.

📄 License

MIT © Kal Aeolian