skills/cloudvoyant/codevoyant/aws/Gen Agent Trust Hub

aws

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches Firecracker releases and microVM kernels from well-known sources, including the official Firecracker GitHub repository and AWS S3 buckets used by the project's CI.
  • [COMMAND_EXECUTION]: Contains numerous shell scripts and Terraform local-exec/remote-exec style instructions for bootstrapping state backends, managing AWS CLI profiles, and deploying application code.
  • [PRIVILEGE_ESCALATION]: Utilizes sudo within setup scripts to install system binaries, manage device permissions for KVM virtualization (/dev/kvm), and configure low-level jailer environments.
  • [PERSISTENCE_MECHANISMS]: Configures system-level cron jobs to serve as cost-management watchdogs, ensuring compute nodes are shut down after periods of inactivity.
  • [CREDENTIALS_UNSAFE]: Provides a credential management script (scripts/get-creds.sh) that interacts with ~/.aws/credentials and AWS Secrets Manager to facilitate CI/CD runner authentication.
  • [PROMPT_INJECTION]: The skill uses instructional language for the agent to load specific recipes based on user tasks, which is standard instructional design and does not attempt to bypass safety filters or override core agent behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 04:54 PM
Security Audit — agent-trust-hub — aws