terraform-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes user-provided Terraform files, which represents an attack surface for indirect prompt injection.\n
- Ingestion points: The scripts
scripts/tf_module_analyzer.pyandscripts/tf_security_scanner.pyread the content of all.tffiles in the user's workspace.\n - Boundary markers: The scripts do not explicitly wrap analyzed content in delimiters to distinguish it from agent instructions when reporting findings to the LLM.\n
- Capability inventory: The included scripts perform local analysis and do not have network access or file-writing permissions.\n
- Sanitization: The analysis is performed via regex-based extraction of metadata rather than direct ingestion of raw text into the prompt, providing a structural boundary but not complete sanitization of user-controlled names and descriptions.
Audit Metadata