Describe images from the terminal using vision LLMs via OpenRouter.
npx skills add https://github.com/Aayush9029/homebrew-tap --skill peek
Please help me install this Agent Skill. Skill: peek Source: https://github.com/Aayush9029/homebrew-tap/tree/main/skills/peek Command: npx skills add https://github.com/Aayush9029/homebrew-tap --skill peek
Describes images from the terminal using vision models, enabling quick understanding and documentation without a GUI.
Describe images from the terminal using vision LLMs via OpenRouter and optionally rename the files based on the descriptions.
High-intent search queries and answers about installing and using this skill.
You can describe images from the terminal by sending files to a vision LLM via OpenRouter. The tool processes single images or batch directories and returns generated text descriptions directly in your command line.
Yes, you can automatically rename image files based on their content. The tool analyzes images using a vision LLM and optionally renames the files according to the generated descriptions to help organize visual assets.
Yes, you need an OpenRouter API key to generate image descriptions. The OPENROUTER_API_KEY environment variable must be set to authenticate requests and access the vision models used for processing.
To batch process a directory of images with a vision model, point the tool at the target directory. It iterates through the images, sends them to OpenRouter, and generates descriptions for each file without a GUI.
The best way to organize image assets without a GUI is using a terminal tool that generates descriptions via vision LLMs. It analyzes image content and optionally renames files based on the resulting text for quick documentation.
magnus919Manage podcast shows, episodes, and analytics via the Transistor.fm v1 REST API.
cgRGMImplement Convex file upload, storage, serving, and deletion workflows.
lhpaulBatch-orchestrate staged AI development workflow items until terminal conditions.
John2143Describe images from local paths or URLs using vision-capable models.
