Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
Dream Loop provides a process for building games and applications where the final visual output matches a generated rendering. It is designed for developers and technical artists who need to translate descriptive prompts into high-fidelity graphical environments.
The workflow begins by creating a dream version of the request via image generation, then iteratively refining the build until the live screenshot aligns with the target image.
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
<PROJECT_PATH>: path to the existing project or the directory for a new project.
<VISUAL_DIRECTION>: description of the desired game, app, or scene.
Image-generation access and vision input, or a user-supplied target image.
Coding subscription tier so the skill can select its documented Plus or Pro workflow.
Use the dream-loop skill for the project at <PROJECT_PATH>. The angle-bracketed value is a placeholder for my project directory. My visual direction is: <VISUAL_DIRECTION>. Determine the appropriate Plus or Pro workflow, inspect whether this is an existing product or a fresh project, and create the realistic target screenshot at `.dream-loop/target.png`. If image generation is unavailable, stop and ask me to provide the target image or connect an image-generation API. For this first task, stop after the target image is ready and explain the next comparison-and-build step.