The Agent Skills Directory

[NO_CODE]: The skill is composed entirely of Markdown documentation and LaTeX templates. It contains no executable scripts (Python, JavaScript, Shell), which significantly limits its ability to perform direct malicious actions.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and assist in editing user-provided educational LaTeX materials. This creates a surface for indirect prompt injection if a processed document contains malicious instructions intended to mislead the agent.
Ingestion points: The agent reads and processes external LaTeX source files provided by the user.
Boundary markers: The skill does not define specific boundary markers for untrusted content.
Capability inventory: The skill itself has no custom executable capabilities; it relies on standard agent file-manipulation tools.
Sanitization: No specific sanitization of external content is defined in the skill instructions.
[DYNAMIC_EXECUTION]: The reference documentation mentions that the LaTeX packages 'minted' and 'pythontex' require the '-shell-escape' flag. While this is standard for these packages to enable code highlighting and embedded Python execution, it is an security configuration that allows the LaTeX compiler to execute shell commands. Users should only use this flag with trusted source files.

didactic-notes