verify-contact-claim
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
npm installandnpm runto execute theverify-contact-claimCLI application located within the local project structure. This is used to perform the verification logic and interface with the CALL-E service. - [SAFE]: The skill incorporates robust privacy and security controls. It automatically scans for and masks sensitive data such as card numbers, one-time codes, and PINs. It also employs a 'trust anchor' logic that refuses to dial any phone number provided by the potentially malicious contact source, requiring the user to provide a number from a physically trusted source. Furthermore, it explicitly instructs the agent to treat call transcripts as untrusted data, effectively mitigating risks associated with indirect prompt injection.
Audit Metadata