program-game-shaders
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and integrate external shader libraries, material parameter files, and engine-imported shader graphs, which represent potential ingestion points for untrusted data.
- Ingestion points: Referenced community shader packs, external material parameter files, and engine-imported assets (SKILL.md).
- Boundary markers: The skill explicitly instructs the agent to treat these sources as unverified and recommends pinning exact versions and performing manual source reviews.
- Capability inventory: The skill allows the use of
write_fileandexecute_commandto compile and implement the shader pipeline. - Sanitization: The instructions mandate enforcing min/max range hints on uniform buffers, clamping artist-facing parameters, and using offline shader validation to prevent malicious or malformed inputs from impacting the system.
Audit Metadata