verified-email

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing externally-sourced digital credentials, which represents an ingestion surface for untrusted data.
  • Ingestion points: Data enters the application context via the credential.credentialJson response and the extracted SD-JWT string as documented in SKILL.md and references/android/identity/digital-credentials/email-verification-implementation.md.
  • Boundary markers: The skill includes multiple explicit warnings and [IMPORTANT] markers stating that client-side parsing is not for validation and that full cryptographic verification must be performed on the server.
  • Capability inventory: The provided Kotlin code examples demonstrate parsing via JSONObject, logging claims with Log.d, and transmitting the raw response to a relying party server.
  • Sanitization: The instructions mandate the use of standard libraries for server-side validation of the issuer, signature, and nonce to prevent spoofing and replay attacks.
  • [SAFE]: All external resource references, including documentation links, standards specifications (W3C, OpenID4VP), and Android dependencies, target trusted organizations or well-known services. The skill adheres to established identity management security patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 01:14 PM
Security Audit — agent-trust-hub — verified-email