The Agent Skills Directory

[SAFE]: The skill performs legitimate infrastructure tasks related to OpenUSD asset management.\n- [SAFE]: All imported modules and external library references (dcc-mcp-core, dcc_mcp_openusd) are consistent with the author's namespace ('dcc-mcp').\n- [SAFE]: No indicators of prompt injection, data exfiltration, or unauthorized command execution were found in the scripts or metadata.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external USD stage files which represent an attack surface, but the risk is managed by vendor libraries.\n
Ingestion points: stage_file parameter in tools.yaml.\n
Boundary markers: None defined for the processed file content.\n
Capability inventory: Allowed tools include Bash, Read, and Write; package_usdz.py performs file writes.\n
Sanitization: Content processing is delegated to the dcc_mcp_openusd runtime library.

openusd-validate