304-frameworks-spring-boot-security
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 standard Maven build commands (
./mvnw compile,mvn compile,./mvnw clean verify) to ensure project stability before and after security modifications. These commands are executed locally within the project environment. - [INDIRECT_PROMPT_INJECTION]: As the skill is designed to read, analyze, and modify existing project source code, it possesses a surface for indirect prompt injection. A malicious actor could embed instructions within the target project's files (e.g., in comments or string literals) that the agent might inadvertently follow during its review workflow.
- Ingestion points: Project source code files,
pom.xml, and configuration properties are read by the agent (SKILL.md, Step 1). - Boundary markers: None explicitly defined to separate project data from instructions.
- Capability inventory: The agent can perform file writes to modify source code and execute shell commands via Maven.
- Sanitization: No specific sanitization or filtering of project content is mentioned before processing.
- [SAFE]: The reference material correctly identifies and warns against common security anti-patterns such as hardcoding secrets (Example 5), using weak password encoders (Example 6), and overly permissive CORS/CSRF configurations (Examples 3 and 4).
Audit Metadata