The Agent Skills Directory

[SAFE]: The skill provides technical instructions for GPU rendering workflows. It does not perform any network operations, access sensitive files, or include instructions that attempt to override agent safety protocols.
[SAFE]: No patterns of prompt injection, data exfiltration, or obfuscation were found. The technical terms (WGSL, TSL, WebGPU) are used correctly within their professional context.
[SAFE]: The skill does not include any dynamic context injection (!command) or remote code execution patterns. References to tools like compare-screenshots are mentioned as part of a development workflow for visual validation.

renderer