The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access behaviors were detected. The skill focuses on providing hardening instructions for MongoDB databases following established security standards.\n- [SAFE]: The skill correctly implements security best practices such as least privilege, encryption at rest and in transit, and auditing, while deferring sensitive tasks like KMS provisioning to infrastructure layers.\n- [SAFE]: All sensitive parameters, such as passwords, certificates, and file paths, are handled using descriptive placeholders (e.g., , ), preventing accidental credential exposure or hardcoding of secrets.\n- [SAFE]: Indirect Prompt Injection Surface: The skill ingests data from external architecture and model files (e.g., architecture/security, data-architecture.md) to generate security configurations. While this is an ingestion point for untrusted data, the risk is negligible as the skill lacks executable capabilities and is specifically designed to enforce restrictive security policies.

mongodb-security-and-data-access-hardening