amazon-brand-registry
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of instructional markdown text and documentation. It does not include any scripts (Python, JavaScript, Shell), binaries, or automation logic that could execute commands on a system.
- [EXTERNAL_DOWNLOADS]: The documentation references the official Amazon Brand Registry domain (brandservices.amazon.com) as a resource for users. There are no automated network requests, script downloads, or remote code execution patterns.
- [DATA_EXPOSURE]: The skill does not attempt to access sensitive files, environment variables, or credentials. It functions by providing structured advice based on user-provided inputs regarding their brand and trademark status.
- [PROMPT_INJECTION]: The instructions do not contain markers for overriding safety filters, extracting system prompts, or instructing the agent to ignore its core programming. The "Role-play" elements are limited to providing a structured multi-step consultation for the user.
- [OBFUSCATION]: A thorough check for Base64 encoding, zero-width characters, and homoglyph attacks revealed no hidden instructions or masked URLs.
Audit Metadata