rtx-remix-modding
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill is authored by a trusted vendor and is intended for use with their official modding toolkit.
- [DATA_EXFILTRATION]: The skill identifies a local configuration file path at %LOCALAPPDATA%\NVIDIA\RTX Remix\mcp.json to discover connection parameters for the Remix MCP server. This is a standard vendor-specific discovery mechanism and does not target sensitive system credentials or user data.
- [INDIRECT_PROMPT_INJECTION]: The skill processes modding data from tool outputs, creating an attack surface for indirect injection. (1) Ingestion points: Data such as prim names and asset paths retrieved via tool calls like remix_get_prim_paths. (2) Boundary markers: Instructions require the agent to relay tool responses and errors verbatim to the user. (3) Capability inventory: The skill can modify project files via remix_save_layer and remix_replace_prim_reference_file_path. (4) Sanitization: The skill is restricted to operations defined by the validated MCP command layer.
- [COMMAND_EXECUTION]: The skill utilizes a set of domain-specific remix_* tools for project management. It includes explicit instructions to avoid manual file editing, which serves as a security measure to prevent unvalidated system commands.
Audit Metadata