safe-action-middleware
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The documentation provides secure coding patterns for the next-safe-action library, including explicit warnings about validating raw client input before use in authorization logic.- [SAFE]: External service integration (e.g., Sentry) follows well-known service patterns for error reporting and monitoring.- [SAFE]: Dependencies mentioned (zod, next-safe-action, Sentry) are industry-standard libraries.- [SAFE]: Analysis of indirect prompt injection surfaces: Ingestion points include clientInput and bindArgsClientInputs (SKILL.md). Boundary markers are provided via explicit documentation warnings. Capability inventory includes database access and logging. Sanitization is demonstrated using Zod for schema validation (auth-patterns.md).
Audit Metadata