The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external HTML and interactive components, which creates a surface for indirect prompt injection. A malicious actor could attempt to influence the agent's behavior via content inside the audited markup. However, the risk is negligible as the skill lacks high-risk capabilities like file writing or network exfiltration.
Ingestion points: Rendered HTML markup and design-system patterns as defined in SKILL.md and references/rule.md.
Boundary markers: None identified; the skill relies on the agent's ability to distinguish between instructions and the data being reviewed.
Capability inventory: The skill's scope is limited to auditing and flagging violations; it does not utilize tools for remote code execution, persistence, or data exfiltration.
Sanitization: No explicit sanitization of the audited content is performed before review.
[DYNAMIC_EXECUTION]: The file references/rule.md includes a Next.js code example that uses dangerouslySetInnerHTML. While this property is a common vector for XSS in web applications, it is presented here solely for educational purposes within documentation and does not pose a threat to the agent's execution environment.

heading-hierarchy