The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted Roblox scripts, which represents a potential indirect prompt injection surface. This is expected behavior for a code refactoring tool.
Ingestion points: The skill ingests script content from the Roblox environment using script_read.
Boundary markers: No explicit delimiters or instructions to ignore embedded code comments/instructions are provided in the workflow.
Capability inventory: The skill utilizes multi_edit for file modifications, script_grep for searching, and start_stop_play for execution testing.
Sanitization: There is no explicit sanitization of the scripts before they are processed by the agent.

roblox-code-maintenance-refactor