All endpoints live on https://api.xona-agent.com and are paid via x402 (Solana USDC micropayments). The agent wallet pays automatically.
Use the x402-fetch.js script in the workspace scripts folder. It handles the 402 payment flow automatically using the agent's Solana wallet.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js <endpoint> [payload_json]
POST example:
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/creative-director '{"idea":"futuristic Solana DeFi dashboard"}'
GET example:
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /token/pumpfun-trending
The script outputs JSON. Parse the response and extract relevant fields (image_url, video_url, banner_url, etc.) to present to the user.
curl -s https://api.xona-agent.com/x402-resources
Research trending design ideas on X and Google, analyze user intent, return refined creative direction.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/creative-director \
'{"idea":"futuristic Solana DeFi dashboard aesthetic","reference_images":["https://example.com/ref.jpg"]}'
| Field | Required | Description |
|---|---|---|
| ------- | ---------- | ------------- |
idea | Yes | Creative idea or design research query |
reference_images | No | Array of reference image URLs |
Response: { success, data: { intent, research, direction } }
The direction array contains generation plans with output_metadata.prompt — feed these into an image generation endpoint.
All image endpoints accept:
| Field | Required | Description |
|---|---|---|
| ------- | ---------- | ------------- |
prompt | Yes | Detailed image description |
aspect_ratio | No | 1:1 (default), 16:9, 9:16, 4:3, 3:4 |
referenceImage | No | Array of reference image URLs |
All return: { success, data: { image_url, image_description, metadata } }
Also accepts style (array of keywords to blend into prompt).
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/designer \
'{"prompt":"a glowing Solana logo floating in space","style":["digital art","futuristic","neon"],"aspect_ratio":"16:9"}'
Artistic, handles text-in-image well. Cheapest option.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/grok-imagine \
'{"prompt":"...","aspect_ratio":"1:1"}'
Balanced quality and cost.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image-model/qwen-image \
'{"prompt":"...","aspect_ratio":"1:1"}'
ByteDance photorealistic model. Best for high-detail, realistic images.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image-model/seedream-4.5 \
'{"prompt":"...","aspect_ratio":"1:1"}'
Fast creative generation. Good default.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/nano-banana \
'{"prompt":"...","aspect_ratio":"1:1"}'
Highest quality.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /image/nano-banana-pro \
'{"prompt":"...","aspect_ratio":"1:1"}'
| Model | Endpoint | Price | Best for |
|---|---|---|---|
| ------- | ---------- | ------- | ---------- |
| Grok Imagine | /image/grok-imagine | $0.04 | Artistic, text-in-image |
| Qwen Image | /image-model/qwen-image | $0.05 | Balanced quality/cost |
| Designer | /image/designer | $0.08 | Style-blended, art-directed |
| Seedream 4.5 | /image-model/seedream-4.5 | $0.08 | Photorealistic |
| Nano Banana | /image/nano-banana | $0.10 | Fast creative, good default |
| Nano Banana Pro | /image/nano-banana-pro | $0.20 | Highest quality |
10-second AI video using Grok Imagine Video. Supports text-to-video and image-to-video.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /video/short-generation \
'{"prompt":"camera slowly orbiting a glowing Solana logo in deep space","aspect_ratio":"16:9"}'
For image-to-video, pass a previously generated image_url as image:
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /video/short-generation \
'{"prompt":"smooth zoom into center","aspect_ratio":"16:9","image":"https://cdn.example.com/img.jpg"}'
| Field | Required | Description |
|---|---|---|
| ------- | ---------- | ------------- |
prompt | Yes | Video description with motion cues |
aspect_ratio | No | Aspect ratio (default: 16:9) |
image | No | Still image URL to animate |
Response: { success, data: { video_url, duration, model, metadata } }
Takes up to 5 minutes. Let the user know it's processing.
Extract latest news from any X account, generate tweet draft and banner.
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /ai/x-news \
'{"x_username":"solana"}'
| Field | Required | Description |
|---|---|---|
| ------- | ---------- | ------------- |
x_username | Yes | X username (with or without @) |
x_persona | No | X username whose style to mimic for tweet draft |
Response: { success, data: { x_username, trending_news: { title, tweet_draft, banner_url } } }
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /token/pumpfun-trending
Response: { success, data: { summary, suggestions, trending_tokens, count } }
node /home/node/.openclaw/workspace/scripts/x402-fetch.js /token/pumpfun-movers
Same response format as trending.
/image/creative-director with the user's ideadirection[].output_metadata.prompt to call an image endpointimage_url to /video/short-generation to animate it/ai/x-news with a target X accounttweet_draft and banner_url/token/pumpfun-trending for current meta/token/pumpfun-movers for biggest moves共 1 个版本