documentary-brutalist-agency
Fail
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: CRITICALINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data such as project descriptions and testimonials to generate site content, creating a vulnerability surface for indirect prompt injection. \n
- Ingestion points: Found in SKILL.md and demo/PROMPT.md, where user-provided photography, quotes, and claims are incorporated into the design process. \n
- Boundary markers: There are no clear delimiters or instructions to treat user data as untrusted content, which could lead to instructions within the data being executed by the agent. \n
- Capability inventory: The skill is limited to generating frontend code (HTML/CSS/JS) and does not possess capabilities for system file access or network exfiltration. \n
- Sanitization: The instructions lack guidance on sanitizing or validating user-provided strings before they are used in code generation.\n- [DYNAMIC_EXECUTION]: The skill ships with a functional demo that includes executable client-side scripts. \n
- Evidence: demo/index.html contains inline JavaScript used to manage the state and accessibility features (ARIA attributes and focus) of a navigation menu.\n- [EXTERNAL_DOWNLOADS]: The skill provides references to external design inspiration and live websites. \n
- Evidence: REFERENCES.md includes links to design templates and portfolios hosted on Framer's official domains.
Recommendations
- CRITICAL: 2 file(s) identified as malware by FileRep - DO NOT USE
Audit Metadata