opencl-runtime

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process kernel source files and build options, which provides an entry point for untrusted data to enter the agent's context.
  • Ingestion points: The skill ingests kernel code via source files and uses command-line arguments for compilation settings.
  • Boundary markers: No explicit boundaries or instructions are provided to the model to ignore embedded commands in the processed data.
  • Capability inventory: The skill utilizes powerful tools including Bash, Write, and Edit across its scripts.
  • Sanitization: The skill does not implement explicit validation or sanitization of the input source files or build options.
  • [DYNAMIC_EXECUTION]: The skill enables the runtime compilation of OpenCL C source code into executable binary code.
  • Evidence: Logic in the skill references API calls such as clBuildProgram and clCreateProgramWithSource for dynamic kernel construction.
  • Note: This behavior is the core intended functionality of the skill and is required for GPU application development.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:37 AM
Security Audit — agent-trust-hub — opencl-runtime