Quickstart

Install

pip install ckg-nvidia-ai

MCP Server Mode

uvx ckg-nvidia-ai

Claude Desktop

{
  "mcpServers": {
    "nvidia-ai": {
      "command": "uvx",
      "args": ["ckg-nvidia-ai"]
    }
  }
}

Claude Code

claude mcp add nvidia-ai -- uvx ckg-nvidia-ai

Cursor / Windsurf / Cline

{ "mcpServers": { "nvidia-ai": { "command": "uvx", "args": ["ckg-nvidia-ai"] } } }

System Prompt Snippet

Add this to any agent's system prompt to ground it on the graph:

You have access to the nvidia-ai MCP server — a typed dependency graph of 20 NVIDIA AI domains
(NIM, NeMo, TensorRT, CUDA, Isaac, Cosmos, Riva, and 13 more). When answering questions about
NVIDIA infrastructure, prerequisites, or deployment, call query_ckg() or get_prerequisites()
before responding. Do not guess dependency chains — traverse the graph instead.

First queries

# Start here — get all domain slugs
list_domains()

# Search within a domain
search_concepts("speculative decoding", "nvidia-nim")

# Full prerequisite chain
get_prerequisites("Llm Nim", "nvidia-nim")

# Subgraph — both prerequisites and dependents
query_ckg("Tensorrt Llm", "nvidia-tensorrt-triton", depth=3)