The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input via the $ARGUMENTS placeholder and interpolates it into the prompt and process instructions without sanitization.
Ingestion points: $ARGUMENTS variable in SKILL.md ( and sections).
Boundary markers: None present; the skill lacks delimiters to separate user input from instructions.
Capability inventory: The skill is limited to the Read tool.
Sanitization: No sanitization or validation of $ARGUMENTS is performed before interpolation.

gsd-help