The Agent Skills Directory

[Prompt Injection] (SAFE): The content consists entirely of LaTeX documentation and pedagogical guidelines. No instructions were found that attempt to override agent behavior or bypass safety filters.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network exfiltration patterns were detected. Code examples use generic placeholder filenames (e.g., 'data.txt', 'write.py').
[Obfuscation] (SAFE): All files are written in clear, human-readable Markdown and LaTeX. No use of Base64, zero-width characters, or other obfuscation techniques was found.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): While the documentation mentions LaTeX packages like 'didactic' and 'minted', there are no scripts that perform unauthorized package installations or execute remote code (e.g., 'curl | bash').
[Privilege Escalation] (SAFE): No commands or instructions related to elevating system privileges were identified.

didactic-notes