td-review-network
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze external data sources including TouchDesigner error logs, Python script content (DATs), and GLSL shader code (DATs). This creates a vulnerability surface where malicious instructions could be embedded within the project files being reviewed.
- Ingestion points: Diagnostic processes in SKILL.md mention reading "script content" and "shader DAT content" using
get_errorsand other inspection tools. - Boundary markers: No specific delimiters or "ignore instructions" warnings are defined for the agent when processing these external data streams.
- Capability inventory: The skill uses domain-specific tools such as
get_errors,list_operators,get_connections,get_parameters,get_operator_info,inspect_values,get_cook_performance,view_operator, andpulse_parameterto analyze and interact with the TouchDesigner environment. - Sanitization: No sanitization or filtering of script or shader content is specified before the agent processes the information.
Audit Metadata