The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration attempts were detected in the skill instructions or implementation references.
[DATA_EXPOSURE]: The skill advocates for security best practices by explicitly instructing the user to secure /actuator/** endpoints with an Admin role and providing examples for clearing Mapped Diagnostic Context (MDC) to prevent sensitive context leakage across threads.
[INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface where untrusted data (e.g., user IDs from request objects) is ingested into logs. It mitigates this risk by providing code examples that use structured logging arguments (kv) and ensuring the MDC is cleared in a finally block to prevent data from one request contaminating another.

spring-boot-observability