The Agent Skills Directory

Indirect Prompt Injection: The skill is designed to ingest and analyze external investigation logs. While this creates a potential surface for instructions embedded within the log data, the skill includes procedural safeguards such as mandatory log entry citations and explicit gap-flagging protocols to ensure the agent remains grounded in the provided evidence.
Modular Skill Reference: The instructions rely on a separate reference skill (internal-investigation) to handle the substantive log-query logic. This modular design is a standard practice for organizing complex agent workflows and does not introduce unusual security risks.
Absence of Sensitive Operations: The skill does not request network access, file system modifications, or the execution of external scripts. It functions entirely through natural language instructions for data analysis and reporting.

investigation-query