The Agent Skills Directory

[SAFE]: The skill is composed strictly of documentation and instructional guidelines. It does not include any scripts, binary files, or automated commands that could be executed by the system.
[SAFE]: The content promotes secure development practices by instructing the agent to prioritize existing project rules, avoid introducing unrequested dependencies, and verify all changes using the project's own tools.
[SAFE]: No evidence of prompt injection, obfuscation, or credential exposure was found. The instructions explicitly warn against making assumptions and require stopping for user input when encountering conflicts or high-risk architectural changes.

codebase-guardrails