An advanced AI generation skill that leverages Agnes AI's models to create, edit, and synthesize high-resolution images and videos seamlessly.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agnes-image-video
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 agnes-image-video 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 Agnes Image and Video Generation Skill?
The Agnes Image and Video Generation Skill is a highly capable integration designed for Openclaw Skills to connect seamlessly with the Agnes AI multimodal API. It empowers AI agents to generate high-quality visual assets using the Agnes-Image-2.0-Flash and Agnes-Video-V2.0 models. Offering unlimited free access to text, image, and video generation endpoints without requiring registration hurdles like credit cards, this skill serves as a cost-effective powerhouse for developers and automated workflows.
With full OpenAI API compatibility, this skill handles advanced generation tasks. It supports classic text-to-image conversion, intelligent image-to-image transformations, multi-image composition, and highly configurable text-to-video or image-to-video rendering. Integrating this within your library of Openclaw Skills unlocks incredible creative capabilities for your AI coding agents and automated workflows.
Agnes Image and Video Generation Skill Use Cases
- Generating high-quality visual content and assets on-the-fly during software development cycles using Openclaw Skills.
- Transforming existing source images into edited versions with specific stylistic modifications (Image-to-Image).
- Creating high-definition videos from textual descriptions or reference images for marketing, demo, or UI prototyping.
- Merging multiple input images into a single cohesive asset (Multi-image composition) programmatically.
- Building custom automated creative pipelines that require high-throughput visual asset generation.
How Agnes Image and Video Generation Skill Works
- Authentication: The skill initializes by validating and injecting the Agnes API Key from environment variables into the HTTP request headers as a Bearer token.
- Image Processing (Synchronous): When generating images, the skill sends a synchronous POST request to the Agnes API hub endpoint (
/v1/images/generations). It passes the selected model, prompt text, target resolution, and tags (such asimg2imgfor reference-based editing) to instantly receive a structured JSON response containing the generated image URL. - Video Processing (Asynchronous): For video generation, the skill initiates a background task via a POST request to
/v1/videos. Since rendering is asynchronous, the skill receives atask_idand enters an automated polling loop. - Polling & Completion: The skill queries the
/v1/videos/{task_id}status endpoint at regular intervals (e.g., every 5 seconds). Once the status transition goes fromqueuedorin_progresstocompleted, it extracts and returns the finalized MP4 URL.
Agnes Image and Video Generation Skill Setup
1. Register and Obtain API Key
Visit the Agnes API Platform to sign up and retrieve your free API key.
2. Configure Environment Variables
Set your retrieved API key as an environment variable so your Openclaw Skills workspace can access it securely:
export AGNES_API_KEY="your-api-key-here"
3. Install Required Dependencies
Ensure you have httpx installed in your Python environment:
pip install httpx
4. Direct Execution Via CLI
You can use the built-in script included in the skill to run generations directly:
# Generate an Image
python scripts/generate.py --mode image --prompt "A cute golden retriever puppy smiling on a sunny lawn" --size 1024x1024
# Generate a Video
python scripts/generate.py --mode video --prompt "A cat walking on the beach at sunset" --height 768 --width 1152 --num_frames 121 --frame_rate 24
Agnes Image and Video Generation Skill Data Schema & Taxonomy
Image Generation Schema
| Parameter | Type | Required | Description |
|---|---|---|---|
model |
string | Yes | Must be "agnes-image-2.0-flash" |
prompt |
string | Yes | Text description for generation |
size |
string | No | Output resolution: "1024x1024" (default), "1024x768", "768x1024" |
tags |
array | No | Set to ["img2img"] for image-to-image modifications |
extra_body.image |
array | Yes (for img2img) | List of input image URLs to process or compose |
Video Generation Schema
| Parameter | Type | Required | Description |
|---|---|---|---|
model |
string | Yes | Must be "agnes-video-v2.0" |
prompt |
string | Yes | Text representation of the video scene |
num_frames |
integer | No | Total frames. Must satisfy 8n+1 and be <= 441 |
frame_rate |
number | No | Target FPS (1 to 60, recommended: 24) |
Valid Frame Count Table
To ensure API compliance, num_frames must follow the 8n + 1 schema. Common valid values include:
- Short Videos: 81, 121, 161
- Medium Videos: 201, 241, 281, 321
- Long Videos: 361, 401, 441
Video Duration Calculation: Duration (seconds) = num_frames / frame_rate (e.g., 121 frames / 24 FPS = ~5.04 seconds).
Agnes Image and Video Generation Skill Advanced Features
- Multi-Image Compositing: Supply multiple source image URLs under
extra_body.imageto direct the model to combine visual elements from separate references into a unified output. - Asynchronous Task Queue & Polling: Built-in automated polling engine that handles server-side video queueing states (
queued,in_progress,completed,failed) and robustly reports rendering progress. - Flexible Image-to-Video Pipelines: Use both reference keyframes and prompts simultaneously to direct precise motion styling and temporal coherence.
- Zero-Cost Scalability: Compatible with the ecosystem of Openclaw Skills, giving agents an unthrottled, zero-cost pipeline for mass media assets without billing limits.
SKILL.md
Loading
Related Openclaw Skills
License Audit
An automated multi-language license compliance scanner powered by Trivy that identifies high-risk dependencies and exports reports in various formats.
AutoDimensionReport
An automated utility to sort supply chain document packages, extract images, run OCR, verify dimension tolerance consistency, and generate comprehensive quality review reports.
Herdsman Skill
Herdsman Skill is an integration package that exposes local LLMs, TTS, ASR, OCR, and image generation capabilities to external agent platforms.
RAGFlow Workbench
An end-to-end RAG platform workbench for Windows to automate Docker verification, admin bootstrapping, knowledge base CRUD, and parsing lifecycles.
Douyin Script Writer Specialist
An intelligent, context-aware Douyin short video script optimizer that adapts to brand identity, extracts guidelines from brief images, and enforces platform compliance.
National Gaokao Counseling Assistant
An advanced Gaokao admission helper skill designed to fetch historical score lines, calculate admission probabilities, and search for universities and majors.
Featured*
