141-java-refactoring-with-modern-features
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
./mvnw compileand./mvnw clean verify. These commands trigger the Maven build lifecycle, which can execute arbitrary code defined in the project's POM file, Maven wrapper, or test suite. If the skill is used on an untrusted or malicious repository, this could lead to local code execution. - [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external, potentially untrusted Java source code.
- Ingestion points: Reads Java source files and project configuration from the user's workspace in
SKILL.mdandreferences/141-java-refactoring-with-modern-features.md. - Boundary markers: The instructions lack explicit boundary markers or directives to ignore instructions that might be embedded in comments or string literals within the code being analyzed.
- Capability inventory: The skill has the capability to write to the file system (refactoring changes) and execute shell commands via Maven.
- Sanitization: There is no evidence of input validation or sanitization of the code content before it is processed by the LLM to identify refactoring tasks.
Audit Metadata