The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions include the use of system commands to manage a local asset bridge. Specifically, it suggests starting a local server using python3 -m http.server 8777 and terminating it with pkill -f "http.server 8777" to provide URLs for the upload_image tool.\n- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface where it reads and processes data from external sources within the Roblox Studio environment. If these scripts or console outputs contain malicious instructions, the agent could potentially act upon them using its available capabilities.\n
Ingestion points: script_read, script_grep, get_console_output, and inspect_instance tools specified in SKILL.md.\n
Boundary markers: No specific delimiters or "ignore instructions" warnings are defined for the ingested data.\n
Capability inventory: The skill possesses powerful capabilities including execute_luau, upload_image, insert_asset, and simulated input tools (user_mouse_input, user_keyboard_input).\n
Sanitization: No sanitization or validation of the ingested content is described before it is processed by the agent.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the retrieval of external content through tools like insert_asset and search_asset, which interact with the Roblox Creator Store to bring external models, images, and other assets into the development environment.

roblox-studio-mcp