Image Prompt
AI image prompt optimizer. Generate and enhance prompts for Midjourney, DALL-E, and Stable Diffusion. Includes prompt enhancement.
BytesAgain2
@ckchzh
What This Skill Does
Generates, enhances, and optimizes image prompts for AI art tools like Midjourney, DALL-E, and Stable Diffusion. Includes a built-in word bank for styles, lighting, composition, and quality modifiers, plus commands for styled generation, prompt enhancement, negative prompts, templates, and a local prompt library.
Replaces manually crafting detailed prompts and searching for style/quality tags by automating prompt generation and enhancement with curated word banks.
When to Use It
- Generate a detailed prompt from a vague subject like 'a fox in a forest'
- Lock a prompt to a specific art style such as watercolor or pixel art
- Upgrade an existing prompt with quality boosters and negative suggestions
- Get a fill-in-the-blank template for portrait, landscape, or product shots
- Translate a Chinese description into an English image prompt
- Save and search a personal library of reusable prompts
Install
$ openclaw skills install @ckchzh/image-promptπ¨ Image Prompt β AI Art Prompt Generator & Optimizer
Turn a vague idea into a detailed, production-ready image prompt. Built-in word banks for styles, lighting, composition, and quality modifiers β just provide a subject and go.
Commands
generate <subject>
Build a full image prompt from a subject. Randomly picks style, lighting, composition, and quality tags from the internal word bank to create a ready-to-use prompt.
image-prompt generate "a fox in a forest"
style <subject> <style>
Generate a prompt locked to a specific art style. Supported styles:
photorealisticβ hyperrealistic photography lookanimeβ Japanese animation aestheticoil-paintingβ classical oil on canvaswatercolorβ soft watercolor washespixel-artβ retro pixel graphics3d-renderβ Blender/Octane 3D stylesketchβ pencil/charcoal drawing
image-prompt style "mountain village" watercolor
enhance <prompt>
Take an existing prompt and upgrade it with quality boosters, detail tags, and a matching negative prompt suggestion.
image-prompt enhance "a cat sitting on a windowsill"
negative
Print a curated list of commonly used negative prompt terms, grouped by category (anatomy, quality, style, artifacts).
image-prompt negative
template <type>
Get a fill-in-the-blank prompt template for a specific genre. Available types:
portraitβ character/person focuslandscapeβ scenery and environmentsproductβ commercial product shotsfoodβ food photographyarchitectureβ buildings and interiorsfantasyβ fantasy/sci-fi illustration
image-prompt template portrait
random
Roll the dice β generates a completely random creative prompt by combining random subjects, styles, lighting, and composition from the word bank.
image-prompt random
translate <chinese_description>
Convert a Chinese-language description into an English image prompt with appropriate tags.
image-prompt translate "δΈεͺη«εε¨ζ¨±θ±ζ δΈ"
save <name> <prompt>
Save a prompt to your local library for later reuse.
image-prompt save hero-shot "a knight standing on a cliff, dramatic lighting, cinematic, 8k"
list
Show all prompts saved in your local library.
image-prompt list
search <keyword>
Search your saved prompts by keyword.
image-prompt search knight
help
Show the help message with all available commands.
version
Print the current version number.
Examples
Basic generation:
$ image-prompt generate "an astronaut"
π¨ Prompt:
an astronaut, digital painting style, volumetric lighting, rule of thirds composition,
intricate details, sharp focus, 8k resolution, trending on artstation
π Negative:
blurry, low quality, deformed, watermark, text, signature
Styled output:
$ image-prompt style "cherry blossom garden" anime
π¨ Prompt:
cherry blossom garden, anime style, cel shading, vibrant colors, soft ambient lighting,
wide angle composition, detailed background, studio ghibli inspired, high quality anime art
π Negative:
photorealistic, 3d render, blurry, low quality, deformed
Enhancing an existing prompt:
$ image-prompt enhance "a dragon flying over a castle"
π¨ Enhanced Prompt:
a dragon flying over a castle, epic scale, dramatic rim lighting, aerial perspective,
intricate scales and details, sharp focus, 8k resolution, cinematic color grading,
volumetric clouds, masterpiece quality, trending on artstation
π Suggested Negative:
blurry, low quality, deformed, watermark, text, bad anatomy, disfigured
Random prompt:
$ image-prompt random
π¨ Random Prompt:
a mechanical owl perched on ancient ruins, steampunk style, golden hour lighting,
centered composition, intricate gears and clockwork, highly detailed, 4k wallpaper,
concept art
π Negative:
blurry, low quality, deformed, watermark, text, signature
Configuration
| Variable | Default | Description |
|---|---|---|
IMAGE_PROMPT_DIR | $HOME/.image-prompt | Base directory for saved prompts and history |
The data directory is created automatically on first run.
Data Storage
All data is kept under $HOME/.image-prompt/ (or the path set via IMAGE_PROMPT_DIR):
~/.image-prompt/
βββ prompts.txt # saved prompts (one per line: name|prompt)
βββ history.log # command usage log
Prompts are stored as plain text β easy to back up, grep, or move between machines.
Powered by BytesAgain | bytesagain.com | hello@bytesagain.com
Top skills in this category
Nano Banana Pro
@steipeteGenerate/edit images with Nano Banana Pro (Gemini 3 Pro Image). Use for image create/modify requests incl. edits. Supports text-to-image + image-to-image; 1K/2K/4K; use --input-image.
AdMapix
@fly0pantsAdMapix raw data layer for ad creatives, apps, rankings, downloads/revenue, and market metadata. Returns structured JSON from the AdMapix API; the calling ag...
YouTube Watcher
@michaelgatharaFetch and read transcripts from YouTube videos. Use when you need to summarize a video, answer questions about its content, or extract information from it.
SuperDesign
@mpociotExpert frontend design guidelines for creating beautiful, modern UIs. Use when building landing pages, dashboards, or any user interface.
Video Frames
@steipeteExtract frames or short clips from videos using ffmpeg.