Dockerfile

FROM python:3.12-slim

# Install system dependencies
RUN apt-get update && apt-get install -y \
    git \
    curl \
    build-essential \
    libopenblas-dev \
    nodejs \
    npm \
    && rm -rf /var/lib/apt/lists/*

# Set working directory
WORKDIR /app

# Install llama-cpp-python with OpenBLAS and Streamlit
RUN pip install llama-cpp-python[server] streamlit

# Download Qwen2.5-Coder-3B-Instruct GGUF Q4_K_M
RUN mkdir -p /app/models
RUN curl -L -o /app/models/qwen2.5-coder-3b-instruct-q4_k_m.gguf https://huggingface.co/Qwen/Qwen2.5-Coder-3B-Instruct-GGUF/resolve/main/qwen2.5-coder-3b-instruct-q4_k_m.gguf

# Install Claude Code globally
RUN npm install -g @anthropic-ai/claude-code

# Create a dummy config.toml for OpenManus if needed, or just for structure
RUN mkdir -p /app/config
RUN echo '[llm]\nmodel = "qwen2.5-coder"\nbase_url = "http://localhost:8001/v1"\napi_key = "not-needed"\nmax_tokens = 4096\ntemperature = 0.0\n' > /app/config/config.toml

# Copy our custom GUI for Claude Code
COPY app_claude_code_gui.py /app/app_claude_code_gui.py

# Create a startup script
RUN echo '#!/bin/bash\n\
# Start llama-cpp-python server in background\n\
python3 -m llama_cpp.server --model /app/models/qwen2.5-coder-3b-instruct-q4_k_m.gguf --host 0.0.0.0 --port 8001 &\n\
\n\
# Wait for server to start\n\
until curl -s http://localhost:8001/v1/models > /dev/null; do\n\
  echo "Waiting for LLM server..."\n\
  sleep 5\n\
done\n\
\n\
# Start Streamlit UI on port 7860\n\
streamlit run app_claude_code_gui.py --server.port 7860 --server.address 0.0.0.0\n\
' > /app/start.sh && chmod +x /app/start.sh

# HF Spaces expect a web service on port 7860
EXPOSE 7860

CMD ["/app/start.sh"]