The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted financial documents (e.g., prospectuses, marketing materials) which present a surface for indirect prompt injection where hidden instructions could influence the agent's behavior during compliance checks. \n
Ingestion points: External financial documents defined in the skill's description and instructions. \n
Boundary markers: The instructions lack explicit delimiters or specific directives for the agent to ignore instructions embedded within the reviewed documents. \n
Capability inventory: The skill utilizes the search-regulations and generate-report tools, which could be leveraged if the agent obeys instructions from within the documents. \n
Sanitization: No input validation or sanitization of the document content is specified.

document-review