aws-infra
Pass
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to execute AWS CLI commands to query and manage cloud infrastructure.
- Evidence:
SKILL.mdandreferences/aws-cli-queries.mdlist various CLI patterns for AWS services including STS, EC2, S3, IAM, and Lambda. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it processes untrusted data from external AWS services.
- Ingestion points: The skill reads CloudWatch logs via
aws logs get-query-resultsand resource descriptions (e.g.,aws ec2 describe-instances) inreferences/aws-cli-queries.md. - Boundary markers: The skill includes high-level instructions to default to read-only queries and require user confirmation for changes, but it lacks structural delimiters to isolate external data from instructions.
- Capability inventory: The agent can perform significant infrastructure changes, such as deleting S3 objects (
aws s3 rm) or updating services (aws ecs update-service), if instructed to do so. - Sanitization: There is no evidence of input validation or sanitization for the data retrieved from AWS APIs before it is incorporated into the agent's context.
Audit Metadata