The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard build and analysis tools including Maven (mvn), Gradle (gradlew), Git, and grep to perform verification tasks. These commands are strictly aligned with the skill's primary purpose of project validation and security scanning.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external data such as source code, test reports, and security scan outputs which could theoretically contain malicious instructions designed to influence agent behavior.
Ingestion points: The agent is instructed to process local Java source code (.java), configuration files (.yml, .properties), and generated analysis reports (JaCoCo, SpotBugs, OWASP Dependency Check).
Boundary markers: Instructions specify targeted file paths (e.g., src/main/) but do not include explicit prompt delimiters or instructions for the agent to ignore content-embedded commands when interpreting code logic.
Capability inventory: The skill possesses capabilities to execute shell commands for building code, running tests, and scanning the filesystem, creating a potential path for actions based on injected instructions.
Sanitization: No explicit sanitization of code content or test output is performed before the agent reviews it for logic or security issues.

springboot-verification