Skip to content

Tools Reference

All tools are read-only. No writes, no side effects.

list_domains()

Returns all 20 NVIDIA AI domains. Start here — domain slugs are required by every other tool.

list_domains()
# → NVIDIA AI CKG — 20 domains:
#   nvidia-nim, nvidia-nemo, nvidia-tensorrt-triton, nvidia-cuda-toolkit ...

search_concepts(query, domain)

Find concepts by keyword within a domain. Use this to get the exact label before calling get_prerequisites() or query_ckg().

search_concepts("speculative decoding", "nvidia-nim")
# → Speculative Decoding [Optimization]
#   Draft Model [Component]
#   KV Cache [Infrastructure]

query_ckg(concept, domain, depth=3)

Traverse the graph from a concept — both prerequisites and dependents. Shows the full connected subgraph up to depth hops in each direction.

query_ckg("FlashAttention-3", "nvidia-cuda-x-libraries", depth=3)
# → Prerequisites: SRAM Tiling → On-Chip Memory → Warp Occupancy → ...
#   Enables: Multi-Head Attention → KV Cache → Speculative Decoding

get_prerequisites(concept, domain)

Full ordered prerequisite chain — everything needed to understand or deploy a concept, traversed upstream until the chain terminates.

get_prerequisites("Isaac Lab", "nvidia-isaac")
# → Isaac Lab → Isaac Sim → USD Composer → Omniverse Kit → ...

ask_nvidia(question, domain="") — v0.4.0+

Natural-language question answered by a local model, grounded on the CKG. Runs entirely locally via Ollama — the model answers only from graph-declared relationships, not parametric memory.

ollama pull qwen2.5:14b   # one-time setup

ask_nvidia("What does TensorRT-LLM require to run on Hopper?")
# → [REQUIRES] CUDA Toolkit → cuBLAS, CUDA Driver API
#   [REQUIRES] FP8 / FP4 Quantization → Hopper SM90 Architecture
#   [ENABLES]  Triton Inference Server → NIM Microservice Runtime
# ---
# Grounded via TensorRT-LLM · nvidia-tensorrt-triton · model: qwen2.5:14b
Env var Default Purpose
NVIDIA_CKG_MODEL qwen2.5:14b Ollama model
NVIDIA_CKG_OLLAMA http://localhost:11434 Ollama host

If Ollama is not running, ask_nvidia() returns raw graph context so the calling agent can still use it.