Prompt
Create a small visual system
Define a coherent palette, type scale, spacing, and component rules for a specific product.
ClaudeChatGPTGeminiDesign
Dave AI
Refine Critique Loop
Use this plugin when the user has an existing Open Design artifact and wants targeted critique, patching, brand tightening, responsive fixes, or quality improvement without starting over.
instructions.md
--- name: refine-critique-loop description: Use this plugin when the user has an existing Open Design artifact and wants targeted critique, patching, brand tightening, responsive fixes, or quality improvement without starting over. license: MIT metadata: author: open-design-spec version: "0.1.0" --- # Refine Critique Loop ## Workflow 1. Read the existing artifact and identify the user's refinement goal. 2. Run a structured critique for hierarchy, fit, accessibility, responsiveness, and artifact-specific quality. 3. Apply the smallest useful patch. 4. Re-run critique and stop when quality converges or the iteration limit is reached. 5. Return a diff summary and what changed. ## Output Contract Patch the existing artifact and produce `refine-summary.md`.--- name: refine-critique-loop description: Use this plugin when the user has an existing Open Design artifact and wants targeted critique, patching, brand tightening, responsive fixes, or quality improvement without starting over. license: MIT metadata: author: open-design-spec version: "0.1.0" --- # Refine Critique Loop ## Workflow 1. Read the existing artifact and identify the user's refinement goal. 2. Run a structured critique for hierarchy, fit, accessibility, responsiveness, and artifact-specific quality. 3. Apply the smallest useful patch. 4. Re-run critique and stop when quality converges or the iteration limit is reached. 5. Return a diff summary and what changed. ## Output Contract Patch the existing artifact and produce `refine-summary.md`.
Platform guidance
Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.