brand-compliance
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [PROMPT_INJECTION]: The instructions do not contain any patterns designed to bypass agent constraints or override system behavior.
- [DATA_EXPOSURE]: The skill accesses local marketing directories (.aiwg/marketing/brand/) but does not attempt to access credentials, SSH keys, or environment secrets.
- [EXTERNAL_DOWNLOADS]: References the W3C accessibility quick reference guide, which is an official and trusted resource.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user content for compliance validation. 1. Ingestion points: User assets such as landing pages and copy. 2. Boundary markers: None identified. 3. Capability inventory: Local file read/write of reports and config; no network or shell execution. 4. Sanitization: None identified. The risk is minimized by the lack of high-privilege capabilities.
Audit Metadata