302-frameworks-spring-boot-rest
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill mandates the execution of shell commands through the Maven build tool (e.g.,
./mvnw compile,mvn clean verify) to validate the state of the user's project before and after modifications. This operational requirement involves the agent interacting with the local shell environment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from the user's project, including Java source code and OpenAPI specifications. Given that the skill also performs command execution, it presents a surface where malicious instructions hidden in project files could potentially influence agent behavior during analysis or build steps.
- Ingestion points: User-provided source code, configuration files, and OpenAPI definitions.
- Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded prompts within the project files.
- Capability inventory: Local shell execution via Maven commands (
mvn,./mvnw). - Sanitization: No validation or sanitization mechanisms are specified for the content of the files being processed.
Audit Metadata