Adds a vision_read tool, a composer switch and a model picker that send an image file to a separate multimodal model and return only its text description, so a session whose model declares text-only input can still read images without switching models. Models are enumerated at runtime with ctx.llm.listModels; openai-completions and anthropic-messages request shapes are both supported.
This plugin has not published an npm package, so the store does not offer an install command yet. Use the repository link below to view the source.