API Documentation

The Tamarind CLI lets you discover tools, submit and monitor protein, nucleic-acid, and small-molecule jobs, and download results - from your terminal, a script, CI, or an AI coding agent such as Claude Code or Codex.

Every tool and parameter available through the Tamarind API is available from the CLI.

Install

Install once with uv or pipx, then run tamarind from anywhere.

1# Requires Python 3.10+. Install with uv:
2uv tool install tamarind-cli
3
4# …or with pipx:
5pipx install tamarind-cli
6
7# …or the bootstrap installer (installs uv if needed):
8curl -fsSL https://app.tamarind.bio/cli/install.sh | sh
9
10# Latest from source:
11uv tool install "git+https://github.com/Tamarind-Bio/tamarind-cli"

Authenticate

Grab your key from the API Key page, then set TAMARIND_API_KEY or run tamarind auth login.

1# Best for scripts, CI, and AI agents:
2export TAMARIND_API_KEY="YOUR_API_KEY_HERE"
3
4# Or store it interactively in ~/.tamarind/config.json:
5tamarind auth login
6tamarind auth status

Quickstart

The end-to-end flow: discover a tool, inspect its schema, validate, submit, and download results.

1# 1. Find a tool
2tamarind tools --function structure-prediction --modality protein
3tamarind tools --search boltz
4
5# 2. Inspect its parameters and grab a runnable example
6tamarind schema boltz
7tamarind schema boltz --example > job.yaml
8
9# 3. Validate, then submit and wait for results
10tamarind validate boltz --input job.yaml
11tamarind submit   boltz --input job.yaml --name my-run --wait --download ./out
12
13# 4. Monitor anytime
14tamarind jobs
15tamarind status my-run
16tamarind results my-run --download ./out

Command reference

Discover

`tamarind tools`	List available tools (filter by --modality, --function, --search)
`tamarind modalities`	List modalities (protein, nucleic acid, small molecule, …)
`tamarind functions`	List tool functions (structure prediction, binder design, …)
`tamarind schema <tool>`	Show a tool's parameters; --example prints a runnable job (for tools that ship one)

Submit

`tamarind validate <tool> --input job.yaml`	Validate inputs without submitting
`tamarind submit <tool> --input job.yaml`	Submit a job (--name, --wait, --download, --skip-validate)
`tamarind batch <tool> --input batch.yaml`	Submit many jobs as one batch

Monitor & results

`tamarind jobs`	List jobs (--status, --batch, --limit, --all)
`tamarind status <jobName>`	Show one job's status
`tamarind wait <jobName>`	Block until a job reaches a terminal state (--timeout)
`tamarind logs <jobName>`	Fetch a job's logs
`tamarind results <jobName> --download ./out`	Download a job's result bundle

Files

`tamarind files list`	List workspace files (--types, --search, --limit)
`tamarind files stats`	Summarize workspace files by type
`tamarind files folders`	List workspace folders
`tamarind files upload <path>`	Upload an input file
`tamarind files delete <path>`	Delete a file (or --folder)

Lifecycle & auth

`tamarind cancel <jobName>`	Stop a running or queued job (--batch for a whole batch)
`tamarind delete <jobName>`	Permanently delete a job and its data (--yes)
`tamarind auth login / status / logout`	Store, check, or clear your API key

Every command accepts --json for machine-readable output (the default when output is piped), --profile to switch accounts, and returns stable exit codes - so it works cleanly in scripts and agents.

Use with AI agents

The CLI is built for coding agents: plain flags, JSON output, and stable exit codes mean an agent can run the whole workflow by calling tamarind directly. For terminal agents (Claude Code, Codex, CI), install the CLI and set TAMARIND_API_KEY. For hosted or chat agents (claude.ai, ChatGPT), connect the Tamarind MCP server, which exposes the same operations.

Related resources

tamarind-cli on GitHub - source, installer, and full README
MCP Server - the hosted option for claude.ai, ChatGPT, and other MCP clients
API Key - get your key and review rate limits

Questions or need help? Contact us at info@tamarind.bio.