The Agent Skills Directory

[SAFE]: The skill uses standard Roblox Studio automation tools (execute_luau, script_read, start_stop_play) to interact with the environment. All operations, including reading script source code for leak diagnosis and spawning server sessions for memory analysis, are directly aligned with the skill's stated purpose.
[SAFE]: No data exfiltration or external network requests were detected. Communication is restricted to internal engine services and JSON encoding for result reporting.
[SAFE]: No evidence of obfuscation, prompt injection, or persistence mechanisms. The skill follows best practices for scene analysis, providing transparent documentation of the data schemas it processes.

scene-analysis