The Agent Skills Directory

[SAFE]: The skill uses Python's standard library exclusively, avoiding the risks associated with unverified or malicious third-party packages.
[SAFE]: Authentication credentials are managed securely in ~/.agentfactory/credentials.json with restricted access permissions (chmod 600), preventing other users on the system from accessing sensitive tokens.
[SAFE]: External network calls are strictly limited to the author's verified infrastructure (panaversity.org) for retrieving lesson content and managing OAuth2 authentication flows.
[SAFE]: Subprocess calls are limited to executing internal skill scripts (auth.py) using the verified local Python interpreter path.
[SAFE]: Prompt instructions in the reference files are focused on teaching methodology and do not contain patterns indicative of prompt injection or safety bypass attempts.
[SAFE]: The skill provides a clear session management and state persistence model using a dedicated directory in the user's home folder, which is standard for personalized CLI tools.

learn-agentfactory