lab-ordering
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill avoids hardcoding sensitive information by using a local configuration file (~/.config/lab_ordering/config.yaml) for identity and funding sources.\n- [SAFE]: The skill requires mandatory human review before any purchase or form submission, providing a strong safeguard against unintended actions.\n- [PROMPT_INJECTION]: The skill processes user-pasted order requests, creating an indirect prompt injection surface. This is mitigated by human-in-the-loop checkpoints and is necessary for the skill's primary function.\n
- Ingestion points: Step 1 of SKILL.md parses user-provided lab order requests.\n
- Boundary markers: No delimiters are used to separate user data from instructions.\n
- Capability inventory: The skill can fill forms on Amazon Business and other sites using browser tools as described in the reference files.\n
- Sanitization: User input is used directly for form filling without specific sanitization or escaping.
Audit Metadata