Kind Installation

This guide installs AI Factory Operations Agent on a disposable local Kind cluster for development and evaluation. It uses an external OpenAI-compatible LLM and enables the local read-only Kubernetes integration. Base Command Manager, Slurm, Grafana, Prometheus, Research Agent, and Hardware Agent integrations are disabled because a fresh Kind cluster does not provide those services.

1. Prerequisites

Install Docker, Kind, Helm 3, and kubectl. Confirm each command is available:

docker version
kind version
helm version
kubectl version --client

Ask your NVIDIA contact to invite the email address you will use with NGC to the AI Factory Operations Agent afoa-release team in the private registry. Accept the invitation while signed in with that account; use a private browser window if another NGC account is already signed in. Then create an NGC personal API key from NGC Setup > API Keys, include the Private Registry service, and confirm that the account can pull the required private registry artifacts.

Obtain an API key, /v1/chat/completions base URL, and model name from the same OpenAI-compatible LLM provider.

2. Create The Cluster

kind create cluster --name mosaic
kubectl config use-context kind-mosaic
kubectl cluster-info

Kind creates the default standard storage class used by AI Factory Operations Agent.

3. Configure Credentials And Dependencies

The NVIDIA API Catalog uses the public endpoint and model below. Replace the endpoint and model together when using OpenAI or another provider.

export EXTERNAL_LLM_BASE_URL='https://integrate.api.nvidia.com/v1'
export EXTERNAL_LLM_MODEL='nvidia/nemotron-3.5-super-vl-120b-a12b'
MOSAIC_CHART=oci://nvcr.io/0948643769302270/afoa-release/mosaic-stack
read -rsp 'NGC API key: ' NGC_API_KEY; echo
read -rsp 'External LLM API key: ' EXTERNAL_LLM_API_KEY; echo

printf '%s' "$NGC_API_KEY" | helm registry login nvcr.io \
  --username '$oauthtoken' \
  --password-stdin

kubectl create namespace aifoa --dry-run=client -o yaml | kubectl apply -f -
kubectl -n aifoa create secret generic mosaic-external-llm \
  --from-literal=apiKey="$EXTERNAL_LLM_API_KEY" \
  --dry-run=client -o yaml | kubectl apply -f -

4. Install AI Factory Operations Agent

# Remove --devel for the latest stable release, or replace it with --version 0.0.1 to pin that release.
helm upgrade --install aifoa "$MOSAIC_CHART" \
  --devel \
  -n aifoa \
  --create-namespace \
  --set global.registryCredentials.create=true \
  --set-string global.registryCredentials.password="$NGC_API_KEY" \
  --set llm.mode=external \
  --set-string llm.external.baseUrl="$EXTERNAL_LLM_BASE_URL" \
  --set-string llm.external.model="$EXTERNAL_LLM_MODEL" \
  --set llm.external.existingSecret=mosaic-external-llm \
  --set modules.kubernetes.enabled=true \
  --set modules.bcm.enabled=false \
  --set modules.slurm.enabled=false \
  --set modules.observability.enabled=false \
  --set modules.grafana.enabled=false \
  --set modules.research.enabled=false \
  --set modules.diagnostics.enabled=false \
  --reset-values \
  --atomic \
  --wait \
  --timeout 12m

unset NGC_API_KEY EXTERNAL_LLM_API_KEY

For upgrades, verification, LLM configuration, module reference, UI access, and headless usage, see the shared Helm guide.

5. Delete The Cluster

kind delete cluster --name mosaic