The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous AWS CLI commands for creating and modifying cloud resources, such as aws lambda create-function, aws lambda update-function-code, and aws s3api put-bucket-notification-configuration. These commands allow the agent to directly mutate the user's AWS environment.
[PRIVILEGE_ESCALATION]: The documentation includes examples of aws lambda add-permission commands that use --principal "*" to grant broad invocation rights to functions and Function URLs. This encourages insecure configurations that could expose the user's infrastructure to unauthorized public access.
[INDIRECT_PROMPT_INJECTION]: The skill defines handler logic that processes untrusted data from external AWS event sources.
Ingestion points: External data enters via the event object in app.py (API Gateway) and sqs_processor.py (SQS records).
Boundary markers: No delimiters or security instructions are provided to help the agent distinguish data from instructions within these events.
Capability inventory: The skill environment allows for file system manipulation (zip), package installation (pip), and infrastructure deployment via AWS CLI.
Sanitization: The provided code snippets perform standard JSON parsing but lack validation or sanitization of the input content before processing.

aws-lambda