The Agent Skills Directory

[SAFE]: The skill is a documentation file defining secure coding standards across six phases: trust boundaries, input/output validation, identity management, secrets/cryptography, robustness, and dependency hygiene.- [SAFE]: While the skill contains code snippets that would normally be flagged as security risks (e.g., hardcoded API keys like sk-proj-..., insecure SQL queries, and dangerous functions like os.system or pickle.loads), these are explicitly documented as "Red flags" to be identified and corrected during code reviews.- [SAFE]: The skill does not perform any network operations, access sensitive files, or execute commands. It functions as a reference library for the agent.

secure-coding