Is andyzwp/image-read safe?
Yes. andyzwp/image-read is safe to install. Oathe's behavioral security audit gave the image-read skill by andyzwp a trust score of 83/100 with 3 findings, none critical or high.
https://clawhub.ai/andyzwp/image-read
Is andyzwp/image-read safe to install?
This skill provides legitimate image analysis functionality using Zhipu AI's GLM-4V-Flash API. While it involves sending user images to an external service, this is clearly documented and represents the core functionality rather than malicious behavior.
What security issues were found in andyzwp/image-read?
Category Scores
Findings (3)
MEDIUM External Image Data Transmission -15 ▶
The skill sends user-provided images to Zhipu AI's GLM-4V-Flash API service for processing. While this is the stated functionality, users should be aware that their image data will be transmitted to and processed by a third-party service.
Evidence: Python script contains: client.chat.completions.create(model='glm-4v', messages=[{content: [image_url]}])
LOW Third-Party API Dependency -10 ▶
The skill requires users to register with and obtain API keys from an external service (bigmodel.cn). This creates a dependency on a third-party service and requires users to share contact information with that service.
Evidence: SKILL.md states: '访问 https://bigmodel.cn/ 注册账号' and '获取API Key:https://bigmodel.cn/console/apikeys'
INFO Non-English Documentation -5 ▶
The skill documentation is primarily in Chinese, which may make it harder for non-Chinese speakers to fully verify the skill's intended behavior and security implications.
Evidence: SKILL.md content is in Chinese language
Should I install andyzwp/image-read?
Oathe's verdict for andyzwp/image-read is SAFE with a trust score of 83/100. Recommendation: Install with caution.