Paper BananaA
Generates publication-ready academic illustrations using the PaperBanana five-agent pipeline (arXiv:2601.23265). Diagram mode runs 5 VLM/image-model agents (Gemini by default; OpenAI, Anthropic, OpenRouter supported) with multimodal in-context learning from curated reference images, venue style packs, and an improve-existing-figure mode. Plot mode generates executable Python matplotlib/seaborn code sized to the venue's column width. Use when the user asks for: research paper figures, academic...
ai-agentspythongo
Skills Directory