The Agent Skills Directory

Indirect Prompt Injection Surface: The skill is designed to ingest and process data from external AWS sources, including cluster configurations, operational logs via logs:FilterLogEvents, and performance metrics from CloudWatch. While processing external data is necessary for its diagnostic purpose, it creates a potential surface for indirect prompt injection if those data sources contain adversarial content.
Evidence: The skill instructions define workflows in SKILL.md to collect data using aws kafka list-cluster-operations-v2 and logs:FilterLogEvents to identify recent issues.
Mitigation: The skill includes strong boundary markers, explicitly instructing the agent that all mutating commands (such as update-broker-storage or reboot-broker) must be presented only as recommendations for a human operator to review and execute, and must not be performed autonomously by the agent.
References to Trusted External Documentation: The skill contains multiple references to official Amazon Web Services (AWS) documentation for best practices, service quotas, and troubleshooting guides. These references provide authoritative context for the agent's analysis and target a well-known service domain.

msk-operations