The Agent Skills Directory

[SAFE]: The skill is a documentation-based resource providing integration patterns and code examples. All identified behaviors align with its primary purpose of AI development assistance.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill promotes secure credential management by demonstrating the use of Spring property placeholders (e.g., ${OPENAI_API_KEY}) and System.getenv() instead of hardcoding secrets.
[EXTERNAL_DOWNLOADS]: All referenced dependencies, such as langchain4j-spring-boot-starter, are well-known, official packages from the LangChain4j project distributed through standard Maven and Gradle repositories.
[PROMPT_INJECTION]: Examples in the documentation utilize @SystemMessage and @UserMessage templates correctly to define AI behavior and constraints, which is a standard practice for mitigating direct prompt injection.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates patterns for ingesting user data (e.g., customer inquiry handling). While this defines an attack surface for the resulting application, the documentation includes remediation guidance such as validating AI-generated outputs and using system messages for behavioral constraints.

langchain4j-spring-boot-integration