Design

What This Skill Does

The Design skill is an adaptive visual intelligence module for OpenClaw. Unlike static design systems, this skill employs a machine learning feedback loop to synthesize your visual preferences over time. It continuously observes your reactions, edits, and creative choices across multiple mediums—including UI layouts, graphic design, video editing, and print media. By maintaining an evolving profile of your 'Aesthetic DNA', the agent ensures that every creative task it performs for you feels native to your personal brand or artistic vision.

Installation

To integrate this skill into your OpenClaw environment, use the following command in your terminal:

clawhub install openclaw/skills/skills/ivangdavila/design

Once installed, the skill generates internal configuration files (dimensions.md and criteria.md) to track your preferences. You can audit these files anytime to see how the agent perceives your taste.

Use Cases

• UI Prototyping: Automatically generate wireframes that align with your preferred color palettes, whitespace usage, and typography hierarchy.
• Content Creation: Produce social media graphics, banners, or video thumbnails that match the visual identity of your previous successful posts.
• Brand Consistency: Ensure all visual assets created by the agent for a specific project adhere to established stylistic rules, reducing the need for iterative corrections.
• Trend Alignment: Adapt to new design trends while filtering out elements that deviate from your core visual identity.

Example Prompts

1. "Analyze my last three project mockups and create a new landing page header that follows the same minimalist aesthetic."
2. "I'm starting a new brand identity for 'EcoSpark'. Based on my previous design preferences, suggest a color palette and font pairing that feels modern yet organic."
3. "Look at the video edits I've rejected recently and tell me if there's a pattern in the color grading or pacing that I should avoid in the future."

Tips & Limitations

• Be Explicit with Feedback: The skill learns best when you provide corrective feedback. If you don't like a design, explain why—the agent will log this to prevent future repetition.
• Consistency is Key: The algorithm requires at least two consistent interactions to establish a rule. Don't expect instant 'tastes' after a single prompt.
• Scope Limits: While powerful, this skill focuses on visual style. It does not replace technical tools like Figma or Premiere Pro; rather, it informs them. It works best as an assistant to augment your creative workflow, not as a complete replacement for human design oversight.