gcp-compute
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous templates for gcloud CLI commands and Terraform configurations to manage cloud infrastructure. These are standard administrative tools for the described use case.
- [INDIRECT_PROMPT_INJECTION]: The skill defines templates that interpolate user-provided variables (e.g., $PROJECT_ID) into shell commands. This represents a potential surface for injection if user input is not validated by the agent platform.
- Ingestion points: Variables in gcloud command snippets and Terraform resource blocks in SKILL.md.
- Boundary markers: Absent; templates use standard shell variable syntax.
- Capability inventory: Execution of gcloud compute subcommands and Terraform provider operations.
- Sanitization: None; the skill serves as a reference guide for command syntax.
Audit Metadata