The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or data exfiltration vectors were identified. The skill performs legitimate shader rendering tasks using local resources.
[COMMAND_EXECUTION]: The skill utilizes a bash script (scripts/render_headless.sh) to invoke the glslViewer binary. This script incorporates defensive programming practices, such as variable quoting, path normalization via realpath, and input validation using regular expressions for the render size, which effectively mitigates common shell injection and path traversal risks.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests and processes shader files, the risk is minimized as the workflow involves the agent translating the content, and the subsequent rendering step is performed by a specialized graphics tool with no network access or broad system permissions.

glslviewer-shader-debug