Initialize a Softlight project and run background designer prompts for prototypes.
npx skills add https://github.com/orianna-ai/claude-plugin --skill softlight
Please help me install this Agent Skill. Skill: softlight Source: https://github.com/orianna-ai/claude-plugin/tree/main/plugins/softlight/skills/softlight Command: npx skills add https://github.com/orianna-ai/claude-plugin --skill softlight
This Skill helps teams initialize a Softlight project that is backed by a long-running agent, enabling iterative design explorations and automated prototype generation.
Run the setup_project script with a short, 3-5 word title to initialize your Softlight project.
High-intent search queries and answers about installing and using this skill.
To start design exploration with an autonomous agent, provide the target application and design problem, then run the setup_project script with a 3-5 word title to initialize the project and dispatch the background designer agent.
Design prototype generation requires two context inputs: the application being changed and the design problem to solve. The agent will not proceed with project initialization until both are explicitly provided.
A background designer agent produces design explorations by running long-running prompts after project initialization, autonomously generating and rendering prototypes on the canvas while reporting progress continuously.
Yes, you can observe design prototype progress directly on the canvas as explorations and prototypes appear. The designer agent runs in the background, continuously updating the canvas and reporting progress without blocking.
After initializing an autonomous design project, you receive a project_id and project_url. The designer agent continues running in the background, outputting iterative design explorations and prototypes on the canvas.
A limitation of running long-running agents for design exploration is that the skill will not proceed with project initialization unless both the target application and the design problem are explicitly confirmed beforehand.
YeaChan05Define problem, scope, and acceptance criteria into a /plan artifact.
aayushsoamGenerate batches of images via the OpenAI Images API with Python.
lukacfManages background shell tasks with monitoring, cancellation, and output handling.
SethyPagnaDesign local HTML artifacts for prototypes, decks, and UI modifications.
