511-frameworks-micronaut-jdbc
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
./mvnw compile,mvn compile,./mvnw clean verify, andmvn clean verify. These are standard Maven build and test commands required for the skill's stated purpose of refactoring and verifying Java code. They are performed within the local project environment. - [INDIRECT_PROMPT_INJECTION]: As a code review and refactoring assistant, the skill ingests and processes external data (project source code) that could theoretically contain malicious instructions embedded in comments or strings.
- Ingestion points: Reads project context and source files during the 'Workflow' steps in
SKILL.md. - Boundary markers: The instructions do not specify explicit delimiters for external code, though it assumes a structured senior-engineer role.
- Capability inventory: The skill can execute shell commands (
mvnw,mvn) via subprocess calls. - Sanitization: There is no explicit sanitization of ingested code comments before processing.
Audit Metadata