tao-setup

TAO Setup

One-time session bootstrap for the TAO skill bank. TAO skills are standalone — each model, data, and platform skill carries its own pinned container image and instructions — but multi-skill workflows chain them (data prep, train, evaluate, deploy). This skill provides the session-level pieces that make that chaining work when skills are installed individually: the discovery flow, the credential conventions, and the host preflight.

When the full skill bank is installed as a plugin from this repository, a SessionStart hook injects this guidance automatically and you do not need to run this skill. When skills were installed one-by-one from a skills catalog, run this skill first.

Quick Start

set -a; source /path/to/.env; set +a   # omit if already exported

# 1. Host preflight — most TAO skills dispatch docker containers on a GPU host.
docker info > /dev/null && echo "OK: docker" || echo "MISSING: docker"
nvidia-smi > /dev/null && echo "OK: GPU" || echo "MISSING: NVIDIA GPU/driver"

# 2. Credential presence check — names only, never print values.
for v in NGC_KEY HF_TOKEN WANDB_API_KEY ACCESS_KEY SECRET_KEY S3_BUCKET_NAME S3_ENDPOINT_URL BREV_API_TOKEN; do
  [ -n "${!v:-}" ] && echo "SET:   $v" || echo "unset: $v"
done

# 3. NGC registry login (needed for nvcr.io image pulls). Key goes over
#    stdin — never as an argv flag, where it lands in the process table.
[ -n "${NGC_KEY:-}" ] && printf '%s' "$NGC_KEY" | docker login nvcr.io -u '$oauthtoken' --password-stdin

If Docker or the NVIDIA host runtime is missing, use the tao-setup-nvidia-gpu-host skill — it checks and (with approval) installs NVIDIA driver 580 or newer, CUDA Toolkit 13.0 or newer, and NVIDIA Container Toolkit 1.19.0 or newer, and can install Docker itself on Debian/RHEL/SUSE-family hosts. These are TAO-wide minimums. If the selected model's references/skill_info.yaml declares runtime_requirements.gpu_host, pass those model-specific minimums to the host setup skill instead.