The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to have the agent ingest and analyze data from various untrusted external sources to generate content ideas.
Ingestion points: The 'Content Ideation Sources' section in SKILL.md instructs the agent to process keyword exports, sales call transcripts, survey responses, and forum content from sites like Reddit and Quora.
Boundary markers: The instructions do not explicitly require the use of boundary markers or delimiters when the agent processes these external inputs.
Capability inventory: While the skill itself is instructional, the agent's overall capabilities (such as web search or file writing) could be manipulated if instructions are embedded within the analyzed content.
Sanitization: The skill lacks specific instructions for the agent to sanitize, filter, or ignore potential instructions found within the data sources being analyzed.

content-strategy