skills/nvidia/skills/nvshmem-docs/Gen Agent Trust Hub

nvshmem-docs

Pass

Audited by Gen Agent Trust Hub on Oct 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a documentation retrieval tool for NVSHMEM, authored by NVIDIA and targeting official NVIDIA resources.
  • [EXTERNAL_DOWNLOADS]: The skill retrieves content from NVIDIA's official domains, including docs.nvidia.com, archive.docs.nvidia.com, and catalog.ngc.nvidia.com. As these are official services of the skill's author, they are considered trusted sources.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes content from external documentation pages, which constitutes a potential attack surface.
  • Ingestion points: Remote documentation pages fetched via a retrieval tool (referenced in SKILL.md and references/topic-map.md).
  • Boundary markers: The instructions require the agent to cite official links and explicitly state the documentation version used, which helps provide context for the retrieved data.
  • Capability inventory: The skill uses a retrieval tool for read-only access to external documentation; it does not perform file writes, system modifications, or command execution.
  • Sanitization: Content ingestion is limited to restricted, official vendor domains, which significantly mitigates the risk of malicious content injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 31, 2026, 05:37 PM