pw-image-generation_skill

This skill guides you through style analysis, prompt design, and per-image confirmation to generate consistent, cost-efficient AI images.
  • TypeScript

19

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 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 plugins-world/pw-skills --skill pw-image-generation

  • SKILL.md21.2 KB

Overview

This skill implements an end-to-end AI image generation and processing workflow focused on cost control and consistent style. It supports text-to-image, image-to-image, batch generation with per-image confirmation, image hosting management, long-image merging, and PPT packaging. The core idea is: analyze style → design prompts → confirm each generation → save outputs to avoid wasted API usage.

How this skill works

The tool reads prompt files from a local prompts/ folder, shows each prompt for user confirmation (yes/skip/quit), then calls the configured image generation API and saves results to an images/ directory. It can upload reference images to an image host and store deletion links, analyze visual style from a URL or local file to produce prompt suggestions, and post-process outputs by merging images vertically or packaging them into a PPTX. Scripts run with Bun/Node and use a simple config/secrets file or environment variables for API keys.

When to use it

  • You need to generate illustrations, article images, or social media visuals from text prompts.
  • You have a reference image and want new images in a similar style (image-to-image).
  • You want to batch-generate a series of style-consistent images but avoid blind bulk consumption of API credits.
  • You need to merge multiple images into a long vertical image for social posts or tutorials.
  • You want to package generated images into a PPT for sharing or presentations.

Best practices

  • Keep prompts clear and consistent: use a single language and include style, subject, scene, atmosphere, and technical details.
  • Generate 1–2 test images first, adjust prompts, then batch-generate remaining items.
  • Use the per-image confirmation to skip low-value prompts and avoid unnecessary API calls.
  • Name prompt files with numeric prefixes (01-, 02-) so outputs sort predictably.
  • Upload high-quality reference images (>=1024×1024) when doing image-to-image and include explicit constraints in the prompt.

Example use cases

  • Create article illustrations: prepare prompts per section, confirm each generation, and export final images to images/ for publishing.
  • Produce a slide deck: generate images for each slide, then run the PPT packaging script to create a 16:9 PPTX with one image per page.
  • Make a long-form social post: generate or collect images, then merge them vertically into a single long image for platforms like WeChat or Instagram guides.
  • Style transfer workflow: upload a reference image to the image host, use the returned URL in prompts, and generate multiple variations that preserve color and composition.

FAQ

The generation script prompts you before each image. Use 'y' to generate, 'n' to skip, and 'q' to quit. Test with 1–2 images before bulk runs.

What if prompt results are inconsistent across images?

Use the style library presets and keep style descriptions identical across prompts. Add more technical detail (e.g., 'photorealistic, high detail, 8k') to reduce variance.

What dependencies are required for merging or packaging outputs?

Merging long images requires ImageMagick installed and in PATH. PPT packaging and generation scripts run with Bun/Node; no extra npm install is required when invoking via npx bun.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational