The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted research data, creating a surface for indirect prompt injection. Ingestion points: Untrusted data enters the context via 'primary external sources' and 'community posts' during the research phase (SKILL.md). Boundary markers: Several safeguards are present, including requirements for explicit operator approval before publication, mandatory evidence receipts, and categorical review by independent parties (SKILL.md). Capability inventory: The skill mentions 'remote writes' and references distribution endpoints like '/publish' and '/article-creator' (SKILL.md). Sanitization: No automated filtering or sanitization of ingested content is specified in the provided instructions.

blog-writing