message-test-designer
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a robust security model through its 'Unified Skill Contract' and 'Routing Policy'. These frameworks ensure that the agent operates under a least-privilege model, where any action resulting in external side effects (e.g., spending, publishing, or modifying accounts) requires explicit user authorization.
- [INDIRECT_PROMPT_INJECTION]: The skill acknowledges the risk of processing data from untrusted sources and includes specific countermeasures.
- Ingestion points: Processes potentially untrusted input such as scraped web text, customer reviews, comments, emails, and third-party data exports as specified in SKILL.md and references/skill-contract.md.
- Boundary markers: Employs explicit instructional boundaries, telling the agent to treat these sources as evidence only and to disregard any instructions embedded within them.
- Capability inventory: The skill's primary function is analysis and artifact generation; it does not request access to dangerous tools or perform arbitrary command execution in its core instructions.
- Sanitization: Relies on strong instructional guardrails to prevent the agent from executing malicious content found in external evidence.
Audit Metadata