The Agent Skills Directory

[SAFE]: No security threats detected. The skill contains extensive documentation and instructional content focusing on game development best practices and platform-specific security (e.g., RemoteEvent validation and DataStore safety).
[SAFE]: The included Python script scripts/validate_skill.py is a utility for checking Markdown link consistency. It is well-documented, uses standard libraries, and performs only read-only operations on the local skill files.
[SAFE]: The skill uses placeholders for sensitive configuration values in templates (e.g., ROBLOX_OPEN_CLOUD_API_KEY=replace_with_local_secret) and explicitly instructs users and the agent to keep secrets out of version control.
[INDIRECT_PROMPT_INJECTION]: The skill provides workflows for researching internet code and Creator Store assets, which are potential ingestion points for untrusted data. However, it includes mandatory safety playbooks (e.g., reference/CREATOR_STORE_ASSETS.md, playbooks/open-source-code-review.md) and a Luau sanitizer template to review and clean imported assets before use.

rbsmithy