The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process a variety of external, potentially untrusted content sources to extract ideation angles. * Ingestion points: Source material includes YouTube transcripts, newsletter content, and web trend findings as defined in SKILL.md (Step 1). * Boundary markers: The instructions do not specify any delimiters or special markers to separate the untrusted source content from the agent's instructions. * Capability inventory: The skill uses the Agent tool to run subagents in parallel and invokes other skills like creator-stack:voice and creator-stack:brand-guidelines on the extracted content. * Sanitization: There are no instructions for sanitizing, filtering, or validating the input data before processing, which could allow malicious instructions embedded in transcripts or newsletters to influence the agent's behavior.

extract-ideas