The Agent Skills Directory

[SAFE]: The skill follows established security best practices for handling credentials. It explicitly instructs the agent to use environment variables instead of hardcoding secrets and to ensure .env files are added to .gitignore to prevent accidental exposure.
[SAFE]: The instruction set includes specific safety gates, such as requiring explicit user consent before handling actual secrets or making changes to non-LaunchDarkly dependencies in subsequent integration steps.
[SAFE]: All external references, package names, and environment variable conventions (e.g., LAUNCHDARKLY_SDK_KEY) are consistent with official LaunchDarkly vendor documentation and infrastructure.

plan