playbooks

home / skills / openclaw / skills / volcengine-ai-image-generation

volcengine-ai-image-generation skill

/skills/cinience/volcengine-ai-image-generation

This skill generates and refines Volcengine AI images by structuring prompts, setting parameters, and returning usable image links.

npx playbooks add skill openclaw/skills --skill volcengine-ai-image-generation

Review the files below or copy the command above to add this skill to your agents.

Files (4)
SKILL.md
757 B
---
name: volcengine-ai-image-generation
description: Image generation workflow on Volcengine AI services. Use when users need text-to-image, style variants, prompt refinement, or deterministic image generation parameters and troubleshooting.
---

# volcengine-ai-image-generation

Generate and iterate images with clear prompt structure and parameter controls.

## Execution Checklist

1. Confirm model/endpoint and output constraints (size, count, style).
2. Normalize prompt into subject, style, scene, lighting, camera terms.
3. Set generation parameters and run request.
4. Return image links/files with prompt and params.

## Prompt Structure

- Subject
- Composition
- Style
- Lighting
- Quality constraints

## References

- `references/sources.md`

Overview

This skill provides a structured workflow for generating and iterating images using Volcengine AI services. It focuses on clear prompt construction, deterministic parameter control, and delivering image outputs with reproducible settings. Use it to create text-to-image results, style variants, and refined prompts for predictable outcomes.

How this skill works

The skill normalizes an input prompt into discrete components (subject, composition, style, lighting, quality) and validates model/endpoint constraints such as output size and count. It sets deterministic generation parameters, issues requests to the chosen Volcengine endpoint, and returns image links or files alongside the full prompt and parameters used. It also supports iterative refinements and basic troubleshooting guidance when results diverge from expectations.

When to use it

  • Generate images from text prompts with controlled parameters (size, seed, steps).
  • Create multiple style variants of a concept for comparison.
  • Refine prompts to improve composition, lighting, or photographic realism.
  • Reproduce an image generation run deterministically across sessions.
  • Troubleshoot unexpected outputs by checking prompt normalization and parameter settings.

Best practices

  • Normalize prompts into subject, composition, style, lighting, and quality before generation.
  • Confirm model/endpoint capabilities and output constraints (max size, channels, count) first.
  • Fix a seed and deterministic parameters when you need reproducible results.
  • Start with a concise prompt and incrementally add constraints to reach the desired result.
  • Store the final prompt and full parameter set with returned images for auditability.

Example use cases

  • Produce a set of marketing images in multiple visual styles (flat, photoreal, cinematic) for A/B testing.
  • Iteratively refine a character illustration by adjusting composition and lighting between runs.
  • Generate deterministic product mockups at exact dimensions using fixed seeds and parameters.
  • Troubleshoot composition issues by isolating and testing each prompt component (subject, camera, lighting).

FAQ

What prompt structure should I use?

Split prompts into subject, composition, style, lighting, and quality constraints to make each element explicit and adjustable.

How do I ensure reproducible outputs?

Set and record a fixed seed and deterministic generation parameters (steps, sampler, guidance) and confirm endpoint supports deterministic runs.