PixVerse AI Image and Video Generator

PixVerse CLI — generate AI videos and images from the command line. Supports PixVerse V6, Veo, Sora, Grok video models; Nano Banana (Gemini), Seedream, Qwen...

This skill appears to do what it says: a CLI that interacts with PixVerse cloud services. Before installing or running it, consider: (1) Authentication uses OAuth device flow and tokens are stored in ~/.pixverse/ (or you can supply PIXVERSE_TOKEN) — treat that token like any other credential. (2) Local files passed to commands (e.g., --image ./file.jpg or --video ./my-video.mp4) are uploaded to PixVerse cloud for processing — do not upload sensitive or private files. (3) The included scripts (update.sh, check-update.sh) perform git pulls and curl to GitHub; run them only if you trust the repository and understand they may modify local files. (4) Minor metadata inconsistency: SKILL.md references PIXVERSE_TOKEN but the registry metadata lists no env vars — this is optional but worth noting. If you need higher assurance, inspect the upstream pixverse npm package code before installing or use npx to run without global install.

Type: OpenClaw Skill Name: pixverse-ai-image-and-video-generator Version: 1.5.0 The skill bundle provides a comprehensive and well-documented interface for the PixVerse AI image and video generation service. It includes detailed capabilities for asset management, authentication via OAuth device flow, and complex generation workflows. The included maintenance scripts (scripts/check-update.sh and scripts/update.sh) are standard utilities for version checking and updating from the official GitHub repository (PixVerseAI/skills). The instructions for the AI agent, including the prompt enhancement logic in capabilities/prompt-enhance.md, are focused on improving output quality while strictly preserving user intent. No evidence of malicious behavior, data exfiltration, or unauthorized execution was found.