Minimax Image Understanding

使用多模态大模型理解图片内容，生成业务含义描述。支持多种模型：(1) MiniMax VLM (2) OpenAI GPT-4V (3) Claude Vision。用于理解截图、图表、文档照片等，生成精准的文字描述。

This skill appears to do what it says (send a local image to a selected multimodal model and return a description), but before installing or using it you should: - Confirm dependencies: ensure the runtime has the 'curl' binary (used by the MiniMax path) and the Python 'requests' package (used for OpenAI/Anthropic). The skill's metadata incorrectly states "no required binaries". - Consider privacy: the script base64-encodes and sends the entire image to remote APIs. Do not use it on images containing sensitive or private data unless you trust the target service and understand its retention policy. - Verify provider endpoints and keys: validate MINIMAX_API_HOST if you set it (default is https://api.minimaxi.com) and never hard-code API keys; supply them via environment variables as instructed. - Review model choice and costs: using OpenAI/Anthropic may incur usage charges and have different input formats/limits — test with non-sensitive images first. If you want stronger assurance, request an updated skill package that explicitly documents runtime dependencies (curl, requests) and includes checks that fail with clear messages when dependencies are missing.

Type: OpenClaw Skill Name: minimax-image-understanding Version: 1.0.0 The skill bundle provides a utility for image understanding using MiniMax, OpenAI, or Anthropic APIs. The script `scripts/understand_image.py` correctly handles API keys via environment variables and transmits image data to the respective service providers as described. No evidence of malicious intent, data exfiltration to unauthorized endpoints, or command injection vulnerabilities was found.