An advanced prompt engineering skill that generates production-grade, highly controlled prompts for Midjourney and other major image models.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install image-prompt-engineer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install image-prompt-engineer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Image Prompt Engineer?
The Image Prompt Engineer is a highly specialized tool designed to assist developers and creators in generating, refining, and translating high-fidelity prompts for generative AI models. While it defaults to Midjourney V7 and V8.1 parameters, the skill seamlessly adapts to other modern image models when requested. By utilizing precise material, lighting, camera, and compositional semantics instead of generic, overused descriptors, it guarantees deterministic, production-ready outputs.
As a powerful addition to your AI development stack, this tool highlights the capabilities of Openclaw Skills by providing your AI coding agents with the logic to act as expert creative directors. It bypasses conversational filler to output ready-to-paste prompts formatted with exact syntactical rules, multi-prompt weightings, and negative exclusions, driving extreme efficiency in generative design workflows.
Image Prompt Engineer Use Cases
- Rapidly prototyping product designs, architecture concepts, and Scandinavian-style interiors.
- Generating clean, flat vector logos and UI wireframes with zero text artifacts or artificial depth.
- Converting loose user concepts into structured, token-optimized Midjourney v7 or v8.1 prompts.
- Generating variation packs and stylistic variants for design exploration.
- Refinishing and updating older prompt syntaxes to match modern AI model parameters.
How Image Prompt Engineer Works
- Model Identification: Detects the target image model. Defaults to Midjourney (stating 'Defaulting to Midjourney' once) unless another model is explicitly defined.
- Creative Detail Inference: Infers missing variables such as lighting, physical materials, camera lenses, and environment details based on the user's high-level goal.
- Prose Structuring: Organizes visual concepts into structured, grammatical descriptive prose, front-loading critical subjects in the first 10 to 30 tokens.
- Syntax Formatting: Appends exact model parameters (such as
--raw,--stylize,--v 7, or--v 8.1) in the correct execution order at the absolute end of the prompt. - Output Generation: Delivers paste-ready code blocks and variation choices accompanied by minimal, high-value rationale.
Image Prompt Engineer Setup
To integrate this prompt engineer workflow into your environment, install the Openclaw Skills repository and configure the skill dependencies:
# Clone the core repository
git clone https://github.com/openclaw/skills.git
cd skills
# Enable the image prompt engineer skill
claw skills enable image-prompt-engineer
Ensure you have the reference file references/midjourney-prompt-guide.md available in your agent's directory to allow the skill to access advanced parameter details, version comparisons, and styling rules.
Image Prompt Engineer Data Schema & Taxonomy
Skill Metadata & Parameters
This skill is registered within the workspace utilizing specific metadata properties to define its execution context:
| Property | Value | Description |
|---|---|---|
| Name | image-prompt-engineer |
Unique identifier for the skill |
| Version | 1.0.2 |
Current semver release of the prompt logic |
| License | MIT |
Open-source licensing terms |
| Tags | image-prompts, midjourney, prompt-engineering, generative-ai |
Discovery tags within the Openclaw Skills registry |
Midjourney Prompt Taxonomy
For Midjourney workflows, the generated prompt strictly follows this anatomical structure:
- Primary Subject & Visual Objective: First 10-30 tokens defining the focus.
- Composition & Camera: Lenses (e.g.,
24mm wide angle), viewpoints (orthographic view). - Materials & Textures: Physical traits bonded with underscores (e.g.,
matte_polycarbonate_housing). - Environment: Background elements and spatial layouts.
- Lighting: Physical lighting constraints (e.g.,
diffuse North-European daylight). - Style & Technical Constraints: Aspect ratios, parameter adjustments.
- Parameters: Final string ending in
--raw --stylize [X] --v [Y].
Image Prompt Engineer Advanced Features
- Omni Reference (V7) Support: Automatically structures image and style weights using
--oref [URL] --ow 100constraints. - Underscore Token Binding: Intelligently binds multi-word materials (such as
matte_polycarbonate_housing) to ensure the model groups them as single visual concepts. - Aesthetic Precision Controls: Implements strict architectural and minimalist rule systems to avoid subjective terms like "beautiful" or "photorealistic".
- Weight-Balanced Negative Prompts: Safely handles negative conditions via
--noand weighted multi-prompts (e.g.,clutter::-0.5) without creating zero-weight errors. - Model Fallback Logic: Detects optimal features between Midjourney V7 and V8.1 depending on whether speed (HD native renders) or Omni features are requested.
SKILL.md
Loading
Related Openclaw Skills
Stalwart Mail Server on Dokploy with Resend SMTP Relay
An automated skill to deploy Stalwart Mail Server on Dokploy with a Resend SMTP relay fallback for blocked port 25 environments.
RTK Token Saver
A CLI proxy skill designed to compress verbose terminal, build, and git outputs to minimize LLM token usage during agent executions.
Refactor Agents Markdown
An automated workflow optimization skill designed to refactor bloated AGENTS.md and CLAUDE.md files into streamlined, high-performance instruction trees using progressive disclosure.
WordPress VPS Install
An automated workflow to provision, secure, and bootstrap WordPress on a fresh VPS using Dokploy and Docker Compose.
Handoff Receiver
An AI agent workflow optimizer designed to safely ingest, validate, and continue development execution from previous session handoffs without losing context.
Google SERP Scholar Cite Search API
A specialized API integration designed to fetch Google Scholar citation data and export links directly into your workflows.
Featured*
