chanjing-one-click-video-creation

What This Skill Does

The chanjing-one-click-video-creation skill acts as a comprehensive end-to-end video production agent within OpenClaw. It automates the entire lifecycle of a short video project, starting from a single topic input to the delivery of a finished MP4 file. The process encompasses generating the video storyboard, drafting the script, producing TTS (Text-to-Speech) audio, creating digital human avatar clips, generating AI-enhanced b-roll footage, and finally composing all assets using ffmpeg.

The skill is built on a modular architecture that relies on strict render rules to ensure high-quality output. It manages complex workflows where specific components like digital humans and AI visuals are synthesized to match the semantic intent of the user's topic. By abstracting the heavy lifting of video editing, asset orchestration, and format synchronization, it allows users to focus on creative strategy rather than technical video production tasks.

Installation

To integrate this skill into your OpenClaw environment, use the following CLI command:

clawhub install openclaw/skills/skills/binkes/chanjing-one-click-video-creation

Ensure that you have ffmpeg and ffprobe installed on your host system as they are critical dependencies for the rendering pipeline. If you are running the skill outside of a standard repository structure, ensure the CHAN_SKILLS_DIR environment variable is correctly configured to point to your base skills directory.

Use Cases

• Automated Content Creation: Quickly generate scripts and video accompaniments for daily news, educational summaries, or social media commentary.
• Marketing & Promotion: Create consistent video content for product launches by inputting a set of key selling points as a topic.
• Digital Persona Engagement: Utilize the integrated avatar technology to produce consistent口播 (spoken-to-camera) content without manual filming.

Example Prompts

1. "一键生成一个关于人工智能对未来工作影响的视频，时长大约一分钟，风格要求专业且有活力。"
2. "根据选题：'如何通过简单运动保持健康'，帮我策划一段短视频，包含数字人讲解和运动场景混剪。"
3. "做一段关于咖啡知识普及的视频，口播人设选择年轻有元气的形象，分镜要快节奏，画面要精美。"

Tips & Limitations

• Avatar Selection: Always use list_figures --source common --json to compare candidate avatars. Do not rely on hardcoded defaults. Select based on the video's tone (e.g., youthful and energetic for general content, mature/authoritative for professional content).
• Render Integrity: All rendering logic strictly adheres to the documentation in render_rules.md. If you notice rendering issues, ensure your workflow configuration in workflow.json explicitly defines parameters rather than relying on system defaults.
• Input Requirements: Avoid vague or placeholder topics. The skill requires meaningful input to generate quality storyboards and scripts; submissions under 5 characters or generic test strings will be rejected.
• Authentication: Ensure your ~/.chanjing/credentials.json is correctly configured; the skill provides hooks for login if credentials are missing.