The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill ingests data from an external source (AWS SNS) and provides it to the agent without boundary markers. * Ingestion points: SKILL.md (via outputs from aws sns list-topics and aws sns get-topic-attributes). * Boundary markers: Absent. * Capability inventory: Executes AWS CLI and jq commands. * Sanitization: None.
[COMMAND_EXECUTION]: The skill executes shell commands using the aws CLI and jq. These commands utilize environment variables such as $SNS_TOPIC_ARN which could be a vector for command injection if populated with unsanitized external input.

aws-sns