The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze external website content (HTML, metadata, and JSON-LD schema) to verify social profile links. This represents a data ingestion surface where malicious instructions could potentially be embedded in third-party site content. However, the skill's capabilities are limited to auditing and providing recommendations, with no high-risk execution paths.
Ingestion points: Website footer, header, About page, and Organization/Person JSON-LD schema found in SKILL.md and references/rule.md.
Boundary markers: None specified in the instructions.
Capability inventory: The skill is scoped to code review, flagging violations, and describing manual/automated verification steps.
Sanitization: Not explicitly defined in the provided files.
[EXTERNAL_DOWNLOADS]: The skill references documentation and standards from trusted and well-known services.
Evidence: References links to frontendchecklist.io, schema.org, and developers.google.com for implementation standards and validation.

social-profiles