Design illustrated portrait marks with AI
Develop illustrated portrait-mark directions and separate square raster candidates, varying palette and composition while preserving a simple visual language.
Design illustrated portrait marks with AI through deliberate art directions and controlled variations. The illustration-logo-skill workflow proposes a face outline, head-and-shoulders portrait or compact half-body composition with a coherent flat palette and one restrained thematic cue.
The deliverable is separate square raster artwork, rather than promised vector files or trademark clearance. Each agreed candidate is generated once and every returned result is delivered. Minor variation does not authorise automatic retries, silent rejection or ranking; a further creative draw follows your request.
What it does
- Establish a simple illustrated portrait language
- Explore distinct palette families
- Vary crop without changing every structural element
- Add a small book, leaf or similar cue
- Inspect legibility at small icon sizes
- Retain labelled candidates with actual dimensions
Install and use
- Load the complete folder and references in an agent with real image-generation and inspection capability. Prepare provider access or credits, a save location, product context and existing colours.
- Ask: “Propose three portrait-mark directions for a reading companion, using warm and botanical palettes, varied crops and one book cue. Generate six separate images after we choose the directions.”
- Agree the direction and quantity unless the request already authorises that batch. Pass a reference to the image tool only when supported and approved.
- Receive every labelled result with its actual dimensions, then inspect it at 32 and 64 pixels. Request a fresh refinement if needed; resampling or automatic ranking must not stand in for the returned artwork.
FAQ
Source
Made by yuluo-yx. Original skill: illustration-logo-skill, published on GitHub under the MIT licence.
View on GitHub