The Agent Skills Directory

[DATA_EXPOSURE_EXFILTRATION]: The skill templates explicitly include requirements for security modeling, encryption, and secrets management. It instructs agents to ensure no secrets are stored in code or environment files, aligning with security best practices. No patterns for data exfiltration or sensitive file access were found.
[OBFUSCATION]: A thorough review of the skill's Markdown files and metadata found no evidence of hidden content, Base64-encoded strings, zero-width characters, or homoglyph attacks.
[PROMPT_INJECTION]: The instructions are designed to guide the agent in generating technical documentation. There are no patterns that attempt to override system instructions, bypass safety guardrails, or reveal internal system prompts.
[COMMAND_EXECUTION]: The skill does not include any scripts, shell commands, or dynamic context injection patterns. It functions entirely through text-based templates and instructional guidelines.
[EXTERNAL_DOWNLOADS]: The skill references a standard Wikipedia page for educational purposes and does not attempt to download or execute external scripts or packages during runtime.

trd