Run folding, docking, and analysis from your terminal, scripts, or CI, with the same job history as the web app.

The ivybloom CLI runs protein folding, docking, and analysis from your terminal, scripts, or CI. It shares the web app's backend and job history, so a job submitted here shows up alongside your web work.

Reach for it when a workflow has to be repeatable. For the platform as a whole, see the platform overview.

Quickstart

Install, sign in, and fold a sequence:

# install (macos/linux)
curl -fsSL https://www.ivybiosciences.com/install | bash

# or via npm (all platforms)
npm install -g @ivybiosciences/ivybloom-cli

# authenticate with your account
ivybloom auth login --browser

# run your first protein folding prediction
ivybloom run esmfold protein_sequence=MKLVFFAEDVGSNKGAIIGLMVGGVVIA

Need detailed installation instructions?

See the installation guide for manual installation, troubleshooting, and CI/CD setup.

Key features

Authentication options

ivybloom auth login --browser      # Browser OAuth (easiest)
ivybloom auth login --api-key KEY  # Direct API key
ivybloom auth login --device       # Device flow (headless)
ivybloom auth link                 # CLI linking

Job submission

ivybloom run esmfold protein_sequence=MKLVFFAEDV...
ivybloom run admet smiles='CCO'
ivybloom orchestrate "Dock caffeine against 1ABC"   # Natural language

Monitor and download

ivybloom jobs list                 # List all jobs
ivybloom jobs follow JOB_ID        # Live progress
ivybloom jobs download JOB_ID      # Download results
ivybloom batch submit screen.yaml  # Batch processing
ivybloom                           # Interactive dashboard (splash → chat)

AI agent integration

ivybloom mcp serve                 # Start MCP server (core toolset; --toolset all for every tool)
ivybloom run esmfold ... --agent   # Structured JSON output