sqldelight-patterns
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for ingesting and persisting data (such as user profiles and posts) into a local SQLite database.
- Ingestion points: Data enters the system via
UserRepositorymethods and is processed by queries inUser.sqandPost.sq. - Boundary markers: While the documentation does not explicitly define LLM-specific boundary markers, the code implementation uses standard SQL parameterized queries (
?). - Capability inventory: The skill enables persistent storage operations through platform-specific drivers (
AndroidSqliteDriverandNativeSqliteDriver). - Sanitization: The implementation strictly follows best practices by using SQLDelight's type-safe generated code and prepared statements, which prevents traditional SQL injection attacks and provides a safe interface for data handling.
Audit Metadata