by:setup

/setup — Environment Discovery and Configuration

Scan the local system and cloud endpoints to detect available tools, GPU resources, and API keys. Write a complete environment profile.

Instructions

Step 0: Read model profile

MODEL_PROFILE=$(cat .by/config.json 2>/dev/null | grep -o '"model_profile"[[:space:]]*:[[:space:]]*"[^"]*"' | grep -o '"[^"]*"$' | tr -d '"' || echo "balanced")

Model lookup for this command:

Agent	quality	balanced	budget
by-environment	sonnet	sonnet	haiku

Step 1: Ensure config directory

mkdir -p .by
if [ ! -f .by/config.json ]; then
  echo '{"model_profile": "balanced"}' > .by/config.json
fi

Step 2: Spawn by-environment agent

Delegate to a by-environment agent (model per profile table above):

Scan the compute environment and produce a complete inventory.

Local tools — check env vars and PATH for installed tools:

• Protenix: $PROTEUS_FOLD_DIR or which protenix
• PXDesign: $PROTEUS_PROT_DIR or which pxdesign
• boltzgen: $PROTEUS_AB_DIR or which boltzgen

GPU — run nvidia-smi if available. Report:

• GPU model, VRAM, driver version, CUDA version
• If no GPU, note "CPU-only — cloud providers required"

Python environment — check for:

• Python version (require 3.10+)
• Key packages: torch, numpy, biopython, boltzgen

Cloud providers — check API key environment variables:

• TAMARIND_API_KEY (Tamarind Bio)
• ADAPTYV_API_KEY (Adaptyv Bio — lab submission)
• Report configured/missing for each (NEVER log key values)

SSH tunnels — check for any active SSH tunnels to GPU nodes

Write complete results to .by/environment.json.

Step 3: Review environment report

Verify the agent produced valid environment.json with all required sections.

Step 4: Select default provider

Based on the environment scan, recommend a compute provider:

1. Tamarind Bio if API key is configured (preferred — free tier)
2. Local GPU if NVIDIA GPU detected with sufficient VRAM (>=16GB)
3. None — warn that no compute provider is available