webgpu-wgsl-uniformity
Pass
Audited by Gen Agent Trust Hub on Aug 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of markdown documentation and reference materials. It does not include any executable scripts, binaries, or configuration files that perform system-level operations.
- [DATA_EXPOSURE_AND_EXFILTRATION]: No patterns of sensitive data access or external network communication were found. Code examples use standard WebGPU APIs in a purely illustrative context.
- [PROMPT_INJECTION]: The instructions do not contain commands intended to bypass safety guidelines, override agent behavior, or extract sensitive system information.
- [METADATA_POISONING]: While the skill author name in the YAML frontmatter differs from the provided author context, there is no evidence of deceptive intent or malicious metadata that would impact the safety of the agent.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze user-provided WGSL shader code, which represents an inherent surface for indirect prompt injection.
- Ingestion points: User-supplied WGSL source code analyzed via instructions in SKILL.md.
- Boundary markers: No explicit instructions are present to mandate that the agent ignore or sanitize instructions embedded in code comments or metadata.
- Capability inventory: The skill relies on the agent's internal coding knowledge; it does not invoke external tools, file-system operations, or network requests.
- Sanitization: No input sanitization or validation logic is defined for the processed source code.
Audit Metadata