springboot-integration-tests
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to run generated tests (
./gradlew test), perform builds (./gradlew build -x test), and manage Git commits (git add,git commit). These operations are consistent with the skill's primary purpose of automating developer workflows. - [PROMPT_INJECTION]: An indirect prompt injection surface is present in Step 0. The skill reads the
languagefield from.spring-config.jsonand uses it to directive the agent's behavior ("switch ALL your responses to that language immediately"). If the configuration file is sourced from untrusted data, it could be used to manipulate agent responses. - Ingestion point:
.spring-config.json(thelanguageproperty). - Boundary markers: None present; the value is directly interpolated into instructions.
- Capability inventory: Shell command execution via Gradle, file system access for writing Java code and configuration.
- Sanitization: No evidence of validation or filtering for the configuration values before they are used in the prompt context.
Audit Metadata