aso-audit
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acknowledges the risk of processing untrusted data from scraped pages, reviews, and emails. It implements a security best practice by instructing the agent to treat this data exclusively as evidence and ignore any embedded commands.
- Ingestion points: Identified in SKILL.md and references/skill-contract.md as scraped text, reviews, and inbox content.
- Boundary markers: The skill contains explicit directives to the agent to treat external content solely as evidence and specifically ignore any embedded instructions.
- Capability inventory: The skill references connectors for marketing and rev-ops platforms via references/connectors.md, but requires explicit authorization for their use.
- Sanitization: Relies on instruction-level sanitization and a Unified Skill Contract that requires authorization for side effects.
- [SAFE]: The skill documentation emphasizes proper secret management by instructing the agent to keep API keys out of source control. It also mandates that all persistent writes and real-world side effects require explicit user authorization.
Audit Metadata