cloudbase-document-database-web-sdk
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates interactions with a CloudBase document database, creating an ingestion point for untrusted data through database queries and realtime listeners. While the skill emphasizes secure coding practices such as security rule configuration, it lacks specific instructions for boundary markers or sanitization to protect the agent's internal prompt context from malicious data stored in the database.
- Ingestion points: Database snapshots and query results retrieved from NoSQL collections (referenced in
crud-operations.mdandrealtime.md). - Boundary markers: No specific delimiters or warnings to ignore embedded instructions are provided for the agent to use when evaluating database content.
- Capability inventory: Full CRUD operations (add, update, remove, get), complex aggregation, and geolocation queries.
- Sanitization: The skill focuses on application-level error handling and security rule enforcement rather than explicit sanitization of data for the agent's internal reasoning.
Audit Metadata