The Agent Skills Directory

[SAFE] (SAFE): No malicious patterns or security risks identified. The skill contains documentation and code examples for a legitimate research library.
[EXTERNAL_DOWNLOADS] (LOW): The documentation suggests installing the transformer-lens package via pip and downloading pre-trained models from Hugging Face. These are standard operations for AI research and development.
[CREDENTIALS_UNSAFE] (INFO): The API reference includes an example of setting a Hugging Face token using a placeholder ('your_token'). This is informative and does not expose actual credentials.

transformer-lens-interpretability