wa-guardrails
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill provides functionality to generate governance steering documents (e.g.,
CLAUDE.md,.cursor/rules/) intended for automatic ingestion by AI agents. This establishes a persistent context for instructing agents on workload-specific standards. - Ingestion points: Governance instructions are stored in files like
CLAUDE.md,.cursor/rules/, and.kiro/steering/within the repository. - Boundary markers: The generated instructions do not utilize specific delimiters to isolate governance rules from potentially untrusted data processed by the agent.
- Capability inventory: The skill is capable of generating configurations that affect CI/CD pipelines (GitHub Actions), AWS Organizations policies (SCPs), and AWS Config evaluations.
- Sanitization: The instructions do not mention specific sanitization or validation logic for the input data used to construct these governance prompts.
Audit Metadata