onboard-a-list

Pass

Audited by Gen Agent Trust Hub on Oct 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external email list data which constitutes an untrusted input surface.
  • Ingestion points: Data is ingested through the ReadListSnapshot.capturePage port described in references/contracts.md, which retrieves contact, preference, activity, and automation facts from external source accounts.
  • Boundary markers: The instructions require that provider-specific terminology be contained within adapters and that all external data be decoded into owned domain schemas.
  • Capability inventory: Across SKILL.md and references/contracts.md, the skill describes capabilities for reading from source account APIs, performing DNS lookups, and writing to destination email service providers.
  • Sanitization: The skill utilizes the Effect library's Schema module to strictly validate and decode all external inputs (e.g., decodeListImportPolicy) before they are processed by the planner or importer.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 30, 2026, 01:30 PM