The Agent Skills Directory

[SAFE]: The skill body consists of procedural instructions for codebase analysis and implementation. It promotes security best practices, such as avoiding inline secret access (e.g., process.env) and using centralized configuration management.
[SAFE]: The instructions explicitly direct the agent to alert the user if existing conventions cause security vulnerabilities or data loss, prioritizing safety over stylistic consistency in those cases.
[SAFE]: The skill uses a structured approach (The Protocol) that minimizes the risk of introducing redundant or unverified dependencies by requiring manifest/lockfile checks and user approval for new additions.
[SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or unauthorized command execution patterns was found in the skill metadata or body.

codebase-conventions