skills/zapabob/skills/tensorrt-llm/Gen Agent Trust Hub

tensorrt-llm

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to pull official Docker images from the nvidia/tensorrt_llm repository and install the tensorrt-llm package via pip. These are legitimate resources provided by a well-known technology vendor for the skill's stated purpose.- [COMMAND_EXECUTION]: The documentation includes standard operational commands for model serving, including starting the trtllm-serve API, managing clusters with ray, and monitoring hardware with nvidia-smi. These actions are consistent with the intended functionality of high-performance inference deployment.- [INDIRECT_PROMPT_INJECTION]: As an inference framework, the skill processes user-supplied natural language prompts. This represents a standard surface for LLM applications and does not include unsafe interpolation or lack common boundary protections in the provided implementation examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 12:08 PM