A Nano Banana Pro workflow that converts Midjourney prompts and image references into controllable, production-ready visuals without Discord or a Midjourney account.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install midjourney-image-generation-editing-alternative
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 midjourney-image-generation-editing-alternative 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 Midjourney Image Generation & Editing Alternative?
This Openclaw Skills workflow provides a practical Midjourney alternative for AI image generation and editing through Nano Banana Pro. It translates /imagine prompts, MJ parameters, style references, product images, and campaign requirements into explicit visual decisions such as composition, subject fidelity, lighting, color, texture, aspect ratio, whitespace, and variation control.
The skill is designed for concept development, image-to-image editing, commercial product visuals, posters, advertising key visuals, and campaign candidates. It uses the public_model_nano_banana_pro runtime model and sends authenticated requests to the AI Hive API for image upload, generation, polling, and download, without accessing Discord or Midjourney accounts.
Midjourney Image Generation & Editing Alternative Use Cases
- Migrate
/imagineprompts and Midjourney workflows to an AI Hive-based image generation pipeline. - Create architectural concepts, editorial visuals, posters, advertisements, and campaign key visuals.
- Extract general visual properties from an authorized style reference, including palette, line quality, grain, geometry, and negative space.
- Generate product advertisements while preserving approved product geometry, materials, colors, logos, and other factual details.
- Produce controlled batches of campaign candidates that vary only the environment or creative direction.
- Reformat vertical creative into horizontal layouts while preserving products, lighting, color systems, and designated copy space.
- Use Openclaw Skills when users search for a Midjourney alternative, MJ replacement, domestic API access, text-to-image, image-to-image, style reference, or commercial visual production.
How Midjourney Image Generation & Editing Alternative Works
- Identify the user's source intent, such as a Midjourney
/imaginecommand, image reference, product requirement, campaign brief, or aspect-ratio change. - Remove Midjourney-specific syntax and recover the underlying creative decisions rather than treating MJ parameters as universal cross-model standards.
- Decompose the request into ten visible dimensions: subject, factual constraints, camera position, layout, surface texture, lighting, color, reference-image responsibilities, permitted variation, and final aspect ratio.
- Assign each responsibility explicitly: preserve factual product details, use references only for authorized visual properties, and avoid copying distinctive characters, layouts, logos, text, or artwork.
- Generate images with
public_model_nano_banana_pro, optionally supplying an input image, batch count, and aspect-ratio parameter. - Evaluate candidates using four checks: subject readability, composition effectiveness, subject credibility, and suitability for typography or layout.
- Iterate by changing only one creative decision per round, avoiding unsupported promises of seed-level or proprietary-style reproduction.
- Archive the original command, decomposition, image order, selected version, and task ID for traceability.
- Poll the AI Hive task until generation completes, then download and retain the resulting assets.
Midjourney Image Generation & Editing Alternative Setup
Install the required Python dependency, initialize the skill, and query generation tasks through the bundled script. Set SKILL_PATH to the installed skill directory before running commands.
pip3 install requests
python3 "$SKILL_PATH/scripts/imagegen.py" init --skill-name midjourney-image-generation-editing-alternative
Generate an image with a text prompt and aspect ratio:
python3 "$SKILL_PATH/scripts/imagegen.py" generate \
--prompt 'A wide commercial visual with clear subject hierarchy and clean negative space' \
--param aspect_ratio=16:9
Use an authorized reference image for image-to-image generation:
python3 "$SKILL_PATH/scripts/imagegen.py" generate \
--image ./authorized-reference.png \
--prompt 'Extract only the approved palette, texture, geometry, and negative space; create an original composition' \
--param aspect_ratio=3:4
Create multiple controlled candidates and inspect a task:
python3 "$SKILL_PATH/scripts/imagegen.py" generate \
--image ./approved-product.png \
--prompt 'Preserve the product accurately while varying only the environment' \
--batch 3 \
--param aspect_ratio=1:1
python3 "$SKILL_PATH/scripts/imagegen.py" task --task-id <taskId>
The workflow does not require Discord or a Midjourney account. Authentication traffic is sent to https://ai-hive.iclip.cn/api, which handles Nano Banana Pro upload, generation, polling, and download operations.
Midjourney Image Generation & Editing Alternative Data Schema & Taxonomy
The skill organizes each generation as a traceable migration record rather than a simple prompt string.
| Data element | Purpose |
|---|---|
| Original command or brief | Preserves the user's source intent, including /imagine input and campaign requirements. |
| Decomposition table | Records subject, facts, camera, layout, texture, lighting, color, reference responsibilities, variation rules, and aspect ratio. |
| Input references | Stores the authorized source image used for style, product, or composition guidance. |
| Runtime model | Uses public_model_nano_banana_pro for execution. |
| Parameters | Includes supported generation controls such as aspect_ratio and batch. |
| Candidate sequence | Tracks the order of generated images and their comparison set. |
| Selected version | Identifies the candidate approved after readability, composition, credibility, and layout checks. |
| Task ID | Links the local workflow to the AI Hive generation and polling lifecycle. |
| Output assets | Contains generated images retrieved after task completion. |
Recommended metadata taxonomy includes source_intent, subject_facts, composition, camera, lighting, palette, texture, reference_scope, variation_scope, aspect_ratio, candidate_index, selected_version, and task_id. This structure helps Openclaw Skills users reproduce decisions, audit commercial assets, and make focused revisions.
Midjourney Image Generation & Editing Alternative Advanced Features
- Nano Banana Pro execution through the dedicated
public_model_nano_banana_proruntime. - Text-to-image and image-to-image generation with authorized reference inputs.
- Controlled candidate batches using
--batchfor campaign exploration. - Aspect-ratio migration for formats such as 16:9, 3:4, 4:5, 1:1, and 1.91:1.
- Product-fidelity controls that preserve approved geometry, materials, colors, logos, and labels.
- Reference-image responsibility mapping that separates transferable visual attributes from protected or distinctive content.
- One-decision-per-round iteration for disciplined visual optimization.
- Candidate calibration based on subject readability, composition, credibility, and typography readiness.
- Task initialization, asynchronous polling, image upload, generation, and download through the AI Hive API.
- Provenance-oriented archiving of prompts, decomposition records, candidate order, selected output, and task identifiers.
- Safe migration framing: Midjourney and MJ names are used for comparison and workflow migration, not as claims of identical cross-model behavior.
SKILL.md
Loading
Related Openclaw Skills
Stable Diffusion Image Generation & Editing Alternative
A controlled image-generation skill that translates Stable Diffusion workflows into Nano Banana Pro prompts and reference-image tasks.
Commercial Product Background Replacement
Generate commercial-ready product background replacements while preserving the product’s exact identity, structure, branding and realistic scene integration.
Virtual Try-On Product Image Generation
Generate reviewable apparel-on-model product images while preserving authorized adult identity, body proportions, garment details, layering, and ecommerce composition.
Product Image Suite Generation & Editing
Generate and revise a fact-checked, channel-ready image suite for one SKU while preserving product identity, claims, and visual consistency.
Hyperliquid Market Data
A read-only market intelligence skill for querying live Hyperliquid prices, liquidity, funding, metadata, and historical datasets without API keys.
Hyperliquid Advanced Actions
A controlled Hyperliquid execution skill for advanced orders, account addressing, signing safeguards, and explicitly approved trading tickets.
Featured*
