311-frameworks-spring-jdbc

Pass

Audited by Gen Agent Trust Hub on Sep 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION

Full Analysis

[COMMAND_EXECUTION]: The skill requires the use of shell commands such as ./mvnw compile and mvn clean verify for project validation.
[INDIRECT_PROMPT_INJECTION]: The skill reviews untrusted Java source code, establishing an indirect prompt injection surface. 1. Ingestion points: Java files and Maven descriptors. 2. Boundary markers: Absent. 3. Capability inventory: Maven command execution. 4. Sanitization: Focuses on SQL injection prevention in the reviewed code.

Audit Metadata

Risk Level

SAFE

Analyzed

Sep 26, 2026, 08:20 PM