contacts-framework
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill enables the ingestion of untrusted data from the iOS Contacts database.
- Ingestion points: Sensitive data is fetched using methods like
unifiedContacts,enumerateContacts, andCNContactPickerViewControllerinSKILL.mdandreferences/change-history.md. - Boundary markers: The instructions do not specify the use of delimiters or warnings to isolate contact data from the agent's instructions.
- Capability inventory: The skill provides read and write capabilities (
CNContactStore,CNSaveRequest) for the Contacts database. - Sanitization: There is no provision for validating or filtering contact fields before they are used in the agent's prompt context.
Audit Metadata