shiny-firestore-mobile
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for building mobile applications that ingest data from external Firebase Firestore collections. Data retrieved at runtime could potentially contain malicious instructions intended to influence an agent's behavior.
- Ingestion points: Data is retrieved through
store.Get<T>,store.Query<T>, and real-time snapshot listeners likeNotifyOnChange<T>andSubscribeChanges<T>. - Boundary markers: The skill does not provide specific instructions to the agent to generate boundary markers or delimiters when processing database content.
- Capability inventory: The generated code performs network operations (Firebase REST and native SDK calls) and local filesystem operations (native persistent cache).
- Sanitization: No explicit sanitization or validation steps for database-retrieved content are mentioned in the code generation rules.
- [SAFE]: The skill includes responsible security guidance, explicitly warning developers never to embed service-account keys in shipped applications and advising the use of server-side packages for such use cases.
- [SAFE]: The documentation identifies a security gap where Firebase Auth tokens are not yet integrated with the native SDK's request auth, advising developers to scope data by collection path to ensure proper isolation until native integration is complete.
- [SAFE]: External dependencies and tools are sourced from well-known and trusted locations, including the official NuGet registry and Google's Firebase infrastructure.
Audit Metadata