The Agent Skills Directory

[SAFE]: The skill is primarily educational, providing a set of best practices and secure implementation patterns for handling sensitive data (OWASP A03). It demonstrates how to properly use standard libraries like Node.js crypto for AES-256-GCM encryption, implement data masking for PII, and configure secure TLS settings.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill explicitly documents how to avoid data exposure. It identifies vulnerable patterns such as logging passwords or including tokens in URLs and provides remediation code like a custom sanitizeForLogging function to prevent accidental leakages.
[CREDENTIALS_UNSAFE]: The skill correctly advises against hardcoded credentials. It provides examples of secure key management using environment variables, AWS KMS, and Google Cloud Secret Manager. While it contains example strings like 'my-secret-key-123', these are clearly marked as // VULNERABLE examples and do not represent actual secrets used by the skill.
[COMMAND_EXECUTION]: No dangerous command execution or shell injection vectors were detected. The skill uses standard API calls for cloud providers and standard filesystem reads for TLS certificates.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution or download unverified packages. It references standard libraries like mongoose, crypto, and official SDKs (@google-cloud/secret-manager, @aws-sdk/client-kms) which are well-known and trusted for their respective purposes.

sensitive-data-protection