AI Photo Editor for Openclaw

Set the required OneKey Gateway credential before making API or CLI requests:

export DEEPNLP_ONEKEY_ROUTER_ACCESS=your_access_key

The key can be registered or managed through the DeepNLP workspace at https://deepnlp.org/workspace/keys. No Node.js or Python package dependencies are required by the skill itself.

For direct API usage, send a JSON request to the agent router:

curl -X POST "https://agent.deepnlp.org/agent_router" \\
  -H "Content-Type: application/json" \\
  -H "X-OneKey: $DEEPNLP_ONEKEY_ROUTER_ACCESS" \\
  -d '{
    "unique_id": "craftsman-agent/craftsman-agent",
    "api_id": "image_generator",
    "data": {
      "prompt": "Create a futuristic AI poster",
      "images": [],
      "card_count": 1,
      "template_id": "photo-editor",
      "asset_size": [{"width": 1080, "height": 1350, "ratio": "4:5"}],
      "mode": "basic",
      "model": "default",
      "styles": ["Editorial grid layout with deep monochromatic tones"]
    }
  }'

For CLI workflows, use the OneKey agent command with a JSON payload:

npx onekey agent craftsman-agent/craftsman-agent image_generator '{"prompt":"Create a futuristic AI poster","images":[],"card_count":1,"template_id":"photo-editor","api_id":"image_generator","asset_size":[{"width":1080,"height":1350,"ratio":"4:5"}],"mode":"basic","model":"default"}'

The full design_config object can be added when precise typography, page layers, audience profiles, styles, and layout constraints are required.

Request fields

Design configuration taxonomy

• channel_profile and audience_profile: Define the target publishing context and audience.
• age_band, language, and profile_version: Store audience and localization metadata.
• styles: Visual direction blocks such as editorial, futuristic, or monochromatic design.
• constraints: Limits for page count, density, print dimensions, or composition.
• brief_inputs: Structured entries with an ID, type, and content. Common types include prompt, style, and constraint.
• text_styles: Typography definitions for brand, headline, content, and footer layers.
• page_templates: Canvas layouts containing background, brand, image, text, headline, content, and CTA layers.

Response and generated files

Supported layer types include background, brand, image, text, headline, and cta. Each page can also include generation status, attempts, image prompts, and content blocks. Exported deliverables are available from the workspace in PNG, PDF, and PPT formats.