2.6k
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill sketch-illustration- _meta.json304 B
- SKILL.md2.6 KB
Overview
This skill generates hand-drawn style illustrations using Imagen 3 (via the ZenMux API) and automatically uploads results to Feishu. It supports three distinct styles—Sketch, Watercolor, and Flat Vector Retro—optimized for diagrams, PPT graphics, tutorials, and product illustrations. The skill defaults to the Sketch style when no style is specified.
How this skill works
You provide a brief description of the scene, characters, layout, and a title; the skill builds a structured prompt that includes a selected style block and negative constraints, then calls Imagen 3 to render the image. After generation the image is saved to the workspace and automatically sent to a Feishu account or open_id. Retry logic is included to handle occasional API 403 responses.
When to use it
- Create minimalist process diagrams or technical flow illustrations for documentation.
- Produce warm, textured PPT illustrations and tutorial images with a watercolor/colored-pencil feel.
- Generate retro-style cover images, course art, or notebook headers with geometric shapes and bold outlines.
- Request quick hand-drawn-style visuals to accompany product feature descriptions or slide decks.
- Need a consistent set of styled illustrations for a multi-slide presentation or course module.
Best practices
- Specify the desired style (Sketch, Watercolor, Flat Vector Retro); default is Sketch if omitted.
- Provide a short centered title and clear content details: characters, key elements, layout, and color tone.
- Include negative constraints (what to avoid) to reduce undesired artifacts and text placement errors.
- Use the Watercolor template for multi-image, cohesive sets; follow the multi-image workflow for consistency.
- If you hit a 403 from the API, retry the request; ensure your ZenMux API key is valid and has Pro+ access.
Example use cases
- A one-slide technical flow illustration in Sketch style for a product demo slide.
- A series of warm Watercolor tutorial images for a training handout.
- A Flat Vector Retro cover image for a course or NotebookLM presentation.
- Product feature callouts and small illustrative assets for a marketing PPT.
- Step-by-step process visuals for documentation or onboarding materials.
FAQ
The skill uses the Sketch (minimal hand-drawn) style by default.
Where are generated images stored and how are they delivered?
Images are saved to the workspace output folder and then uploaded and sent to the specified Feishu open_id.
Which model and API access are required?
Imagen 3 via the ZenMux integration is used; Pro+ access on the ZenMux provider may be required to use the configured model.