The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest the "current topic of conversation" to generate visual representations. While this constitutes an attack surface for processing untrusted data, the skill's capabilities are limited to generating static visualizations (Markdown, Mermaid diagrams) and writing HTML artifacts to a localized .humanlayer/tasks/ directory. No high-risk capabilities such as arbitrary command execution, network exfiltration, or sensitive file access were identified.
[SAFE]: The instructions focus entirely on visualization techniques using standard Markdown, Mermaid, and a specific HTML artifact tag. The skill does not include external dependencies, remote code execution patterns, or attempts to access sensitive system resources.

show-me