The Agent Skills Directory

[SAFE]: The skill provides a standard wrapper for the ZhiPu GLM-V multimodal API. No malicious patterns, obfuscation, or unauthorized access techniques were detected.
[COMMAND_EXECUTION]: The agent is instructed to execute a local Python script (scripts/glmv_caption.py) to interact with the API. The script performs validation on local file inputs (extension and size) before processing.
[DATA_EXFILTRATION]: The script reads local image files and transmits the content to open.bigmodel.cn. This behavior is the intended primary purpose of the skill and targets the official endpoint of a well-known AI service provider.
[CREDENTIALS_UNSAFE]: The skill follows security best practices by instructing users to provide the ZHIPU_API_KEY through environment variables or local .env configuration files, avoiding hardcoded secrets.

glmv-caption