Image Quality Agent

---
name: image-quality-agent
description: "Use when the user wants to create or edit a still image and the request does not match any domain-specific intent such as paid ad creative, marketplace product photo, styled portrait, ID document, poster/cover, interior design, brand identity, IP character, multi-image composite, old-photo restoration, or YouTube thumbnail. Improves vague or generic requests into concrete, model-ready prompts with deliberate composition, color harmony, lighting, lens, material, and style. Do not use for video, multi-image composites, identity-preserving portrait edits, document-spec photos, or requests that clearly belong to a domain skill."
homepage: https://deepkolor.ai
author: DeepKolor
license: MIT
tags: [image-generation, general, prompt-engineering]
version: lite
---

# Image Quality Agent

> **Lite edition** — routing and core rules for agent discovery. Full prompt templates, platform playbooks, and production workflows run on [DeepKolor](https://deepkolor.ai).

General still images when no domain skill fits — turn vague requests into concrete, polished prompts.

## When to use

See `description` in frontmatter — the agent loads it at startup for routing.

## Workflow

- Route to a domain skill when intent matches (ads, ecommerce, portrait, poster, etc.).
- Layer subject, composition, lighting, lens, material, and style in the prompt.
- Fold exclusions into the prompt body (no separate negative-prompt field).

## Defaults

resolution=1K; ratio by subject

## Tool: `create_image_task`

```ts
{
  prompt: string,
  scene: "text-to-image" | "image-to-image",
  aspectRatio?: string,
  imageUrls?: string[],
  resolution?: "1K"
}
```

Available when your agent runtime exposes DeepKolor generation tools. Otherwise use this skill as a prompt-engineering guide and adapt tool names to your stack.

## Boundaries

- Fallback for general images — defer to domain skills when scope is clear.
- No invented copyrighted characters, real people, or product facts.

## Output

After generation, reply briefly: what was created, key parameters, and one useful next step. Return a full prompt only when the user explicitly asks.

## Full version

Production-grade references (SOPs, templates, platform rules) ship inside [DeepKolor](https://deepkolor.ai) — not in this public lite package.