terraform
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill inherently processes untrusted data from external codebases as part of its infrastructure review functionality.
- Ingestion points: The skill reads several types of untrusted files from the user's workspace, including
.tffiles,.tfvars, lockfiles, CI/CD configuration files, and Terraform plan JSON. - Boundary markers: There are no explicit instructions to use delimiters or boundary markers when reading or processing the content of these untrusted files to separate data from instructions.
- Capability inventory: The agent is instructed to use subprocess calls such as
terraform init,terraform validate,tflint,grep, andawkto process these files, creating a potential path for data to influence command execution or analysis results. - Sanitization: The skill includes a proactive safety rule to 'Redact first'—instructing the agent to identify and mask credentials found in the source files before recording findings, which mitigates the risk of secret exfiltration.
Audit Metadata