deploy
Warn
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- COMMAND_EXECUTION (MEDIUM): The skill executes multiple shell commands across several tools (
pnpm,terraform,git,aws). This represents a significant security surface as the agent is granted the ability to modify local files, execute build scripts, and alter cloud infrastructure. - DATA_EXFILTRATION (LOW): The skill reads infrastructure metadata via
terraform outputandaws cloudfront list-distributions. While used legitimately for deployment, these commands could be repurposed to expose internal configuration or infrastructure IDs. - INDIRECT_PROMPT_INJECTION (LOW):
- Ingestion points: The skill reads and acts upon output from
git status,terraform plan, andawsCLI queries. - Boundary markers: None. The agent processes raw CLI output to make deployment decisions (e.g., confirming a Terraform plan).
- Capability inventory: Full filesystem access for builds, infrastructure-as-code execution (
terraform apply), and S3 bucket synchronization. - Sanitization: No explicit sanitization or validation of the CLI output is defined before the agent uses the data in subsequent steps.
Audit Metadata