chanjing-ai-creation

What This Skill Does

The chanjing-ai-creation skill serves as a comprehensive gateway for integrating Chanjing's advanced AI generation capabilities into your OpenClaw workflow. It allows users to orchestrate the lifecycle of AI-generated content, specifically focusing on text-to-image and image-to-video generation tasks. This skill manages the entire pipeline: from task submission, tracking status, and polling for results, to the selective local download of generated assets. It acts as an abstraction layer over the Chanjing open API, utilizing a standard workflow that ensures tasks are queued correctly and results are retrieved only when requested by the user. By centralizing operations under a simple CLI-based utility, it enables users to leverage powerful models like Seedream 3.0 and Kling v2.1 without needing to write custom API clients or handle complex authentication headers manually.

Installation

To integrate this capability into your OpenClaw agent, execute the following command in your terminal: clawhub install openclaw/skills/skills/binkes/chanjing-ai-creation

Ensure that you have completed the initial configuration for chanjing-credentials-guard before running this skill, as it requires valid AK/SK credentials stored in ~/.chanjing/credentials.json. If you are not authenticated, the script will automatically trigger the login page prompt.

Use Cases

This skill is designed for creative professionals and automation engineers who need to integrate AI generation into their pipelines. Common use cases include generating high-quality marketing assets, creating dynamic video content from static image references, and building automated batch processing workflows for media generation. Whether you are generating a set of images for a design project or creating professional-grade video clips from image inputs, this skill provides the structure required to manage these asynchronous tasks efficiently.

Example Prompts

1. "Use the Seedream 3.0 model to generate a 16:9 cinematic image of a rainy cyberpunk city with neon lights."
2. "Take this image URL [https://example.com/start.jpg] and use the Kling v2.1 model to create a 5-second video of the character turning around, then download it to my local folder."
3. "List my last 5 AI creation tasks and check if the latest one is finished yet."

Tips & Limitations

• Explicit Downloading: This skill is built with a 'safety-first' approach regarding file systems. It will not download generated files automatically. You must explicitly request a download for the files to be saved to outputs/ai-creation/.
• Task Lifecycle: Always use poll_task for long-running processes. Do not attempt to manually trigger multiple submissions if you are waiting for a specific output URL.
• Advanced Control: For models requiring unique, non-standard parameters, utilize the --body-file argument to pass a custom JSON payload, ensuring maximum flexibility.
• Preconditions: Ensure network connectivity to https://open-api.chanjing.cc is stable, as the skill relies on continuous communication with the remote server for task status updates.