aoment-image-video（图像视频）

aoment-image-video

AI generation service powered by Aoment AI. This skill exposes dedicated skill APIs and scripts for:

• text-to-image
• image-to-image
• video generation
• image HD repair/upscale

Download the latest package: https://www.aoment.com/downloads/aoment-image-video-skill.zip

Current Version

version: 1.1.0

Auto-Update Policy

Before each invocation, the Python scripts read the version field above and query:

GET https://www.aoment.com/api/skills/aoment-image-video/version

If the local version is behind the remote version, the script exits with:

{
  "success": false,
  "error": "update_required",
  "current_version": "1.0.0",
  "latest_version": "1.1.0",
  "message": "Skill version is outdated..."
}

If the version check fails because of a network problem, the script continues normally.

Quick Start

# 1. Register an Agent account and get your API Key
uv run {baseDir}/scripts/aoment_register.py --nickname "MyBot"

# 2. Generate an image with the default N2-Fast model
uv run {baseDir}/scripts/aoment_image_video.py -k <your-api-key> -t text-to-image -p "a cute cat playing in a garden"

# 3. Repair/upscale an image
uv run {baseDir}/scripts/aoment_hd_repair.py -k <your-api-key> --image ./input.png --resolution 4K

# 4. Check remaining quota
uv run {baseDir}/scripts/aoment_quota.py -k <your-api-key>

Authentication

This skill requires an Agent API Key via:

Authorization: Bearer <api_key>

The API Key format is aoment_ followed by 32 hex characters.

Get your API Key - Agent Registration

AI Agent Bots can register directly via CLI. No web login is required:

uv run {baseDir}/scripts/aoment_register.py --nickname "MyBot"

Or register via API directly:

curl -X POST https://www.aoment.com/api/skills/aoment-image-video/register-agent \
  -H "Content-Type: application/json" \
  -d '{"nickname": "MyBot"}'

Registration response:

{
  "success": true,
  "data": {
    "username": "agent_a1b2c3d4...",
    "nickname": "MyBot",
    "api_key": "aoment_a3f8e1b2c4d6e8f0a1b3c5d7e9f0a1b2"
  }
}

Save the returned api_key; it is used for all subsequent skill calls.

Tools

Available Models

Use the model ID exactly as shown in the --model parameter.

Image Models

Tip: N-series models use Nano Banana Pro, N-Fast-series models use Nano Banana 2, and O-series models use GPT Image 2.

Video Models

text-to-image

Generate images from a text prompt. The default model is image-n2-fast (N2-Fast).

uv run {baseDir}/scripts/aoment_image_video.py \
  --api-key <your-api-key> \
  --tool-type text-to-image \
  --prompt "a cinematic robot painter in a bright studio" \
  --aspect-ratio 1:1 \
  --image-size 1K

image-to-image

Generate a new image from a prompt and a reference image. The reference image can be a URL or base64 image data.

uv run {baseDir}/scripts/aoment_image_video.py \
  --api-key <your-api-key> \
  --tool-type image-to-image \
  --prompt "change the background to a beach" \
  --reference-image "https://example.com/photo.jpg"

video-generation

Generate a video from a prompt. The default and currently supported video model is video-v1.

uv run {baseDir}/scripts/aoment_image_video.py \
  --api-key <your-api-key> \
  --tool-type video-generation \
  --prompt "sunset beach timelapse" \
  --orientation landscape \
  --resolution standard

hd-repair

Repair and upscale an image. This is provided by a separate script:

uv run {baseDir}/scripts/aoment_hd_repair.py \
  --api-key <your-api-key> \
  --image ./input.png \
  --resolution 4K

Quota

Query remaining daily generation quota:

uv run {baseDir}/scripts/aoment_quota.py --api-key <your-api-key>

If your daily quota is used up and you need more, join the community:

• Discord: https://discord.gg/3BMzRd7bJx
• QQ Group: 474397920 (Join via link)

Response Format

All scripts print JSON to stdout.

Successful text-to-image or image-to-image:

{
  "success": true,
  "tool_type": "text-to-image",
  "data": {
    "image_url": "https://cos.example.com/result.jpg?..."
  }
}

Successful video generation:

{
  "success": true,
  "tool_type": "video-generation",
  "data": {
    "video_url": "https://cos.example.com/result.mp4?..."
  }
}

Successful HD repair:

{
  "success": true,
  "tool_type": "hd-repair",
  "data": {
    "image_url": "https://cos.example.com/hd-repair-result.png?..."
  }
}

Successful quota query:

{
  "success": true,
  "data": {
    "remaining": 12,
    "quota": 15,
    "used": 3
  }
}

Error response:

{
  "success": false,
  "error": "error description"
}

Downloading Results

Returned image_url and video_url values are pre-signed COS URLs. Use the complete URL exactly as returned, including all query parameters. Do not strip the query string.

Example:

uv run {baseDir}/scripts/aoment_image_video.py \
  -k <your-api-key> \
  -t text-to-image \
  -p "prompt" > result.json

curl -L -o output.jpg "$(python3 -c "import json; print(json.load(open('result.json'))['data']['image_url'])")"

Troubleshooting

1. If a request fails because of content policy, revise the prompt or reference image and retry.
2. If the script returns update_required, download and install the latest skill package.
3. If a generated URL cannot be opened, make sure your application preserves the full signed URL.
4. For help, join the Discord or QQ community listed above.