The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided story text, which constitutes an inherent surface for indirect prompt injection where instructions could be embedded within the input data.
Ingestion points: Untrusted story text provided as input for analysis as described in SKILL.md.
Boundary markers: The skill includes an explicit 'Constraints' section in SKILL.md that instructs the agent to stick strictly to the provided text and not self-create.
Capability inventory: The skill is restricted to using only the Read tool according to the YAML frontmatter.
Sanitization: There are no specific sanitization or escaping mechanisms implemented for the input text.

character-relationships