web-imagegen-prompting git:20260905.f7824baA

Immutable. This exact content is served forever at /api/v1/blob/457ae6b97f81ca3a.

---
name: web-imagegen-prompting
description: "Write image prompts with crop/placement intent."
---

# Web Imagegen Prompting

Use this skill to make generated imagery usable as website media instead of vague atmospheric filler.

## Workflow

1. Inspect the project, media assets, framework image/video APIs, responsive breakpoints, accessibility requirements, and verification commands.
2. Read [Web Imagegen Prompting Guide](references/web-imagegen-prompting-guide.md) for detailed direction before changing visuals or code.
3. Define the media's job: proof, desire, navigation, instruction, atmosphere, conversion, or accessibility.
4. Implement with stable dimensions, responsive crops, useful fallbacks, and real content hierarchy.
5. Inspect rendered desktop and mobile states, then refine crop, contrast, loading, labels, captions, and polish.

## Scoped execution

Apply the [shared web contract](../website-operating-rules/references/scoped-web-contract.md) once when execution crosses phases; reuse it if already read. Load only the specialist guidance relevant to the requested surface and deliverable.

## Scope and evidence

Prompt-only requests end with usable prompts and crop/placement constraints. Generation and image editing use the available image tool only when in scope; website integration and rendered crop QA apply only to actual implemented assets.