generated: '2026-08-02'
method: searched
source: https://github.com/terraytherapeutics
notes: >-
  Terray Therapeutics publishes no public API and therefore ships no API client
  SDKs. What it does publish are first-party open-source research code libraries
  on GitHub, installed from source (pip install .) rather than from a package
  registry. Probed npm, PyPI, Maven Central, NuGet, pkg.go.dev, RubyGems,
  Packagist and crates.io on 2026-08-02: no first-party Terray package is
  published to any of them (the PyPI project `coati` is an unrelated PowerPoint
  reporting tool by Intelimetrica). No `type: SDKs` pointer is emitted because
  none of these are API client libraries.
registries_probed:
- registry: npm
  result: no first-party package
- registry: pypi
  result: no first-party package
- registry: maven-central
  result: no first-party package
- registry: nuget
  result: no first-party package
- registry: pkg.go.dev
  result: no first-party package
- registry: rubygems
  result: no first-party package
- registry: packagist
  result: no first-party package
- registry: crates.io
  result: no first-party package
packages:
- language: python
  registry: github
  name: coati
  url: https://github.com/terraytherapeutics/COATI
  install: pip install git+https://github.com/terraytherapeutics/COATI.git
  license: Apache-2.0
  official: true
  kind: research-library
  description: >-
    COATI — multi-modal contrastive pre-training for representing and traversing
    chemical space. The chemistry foundation model behind the EMMI platform.
- language: python
  registry: github
  name: coatiLDM
  url: https://github.com/terraytherapeutics/COATI-LDM
  install: pip install git+https://github.com/terraytherapeutics/COATI-LDM.git
  official: true
  kind: research-library
  description: >-
    COATI-LDM — latent diffusion for conditional generation of molecules;
    companion code to the bioRxiv preprint 10.1101/2024.08.22.609169.
- language: python
  registry: github
  name: terramax
  url: https://github.com/terraytherapeutics/terramax
  install: pip install git+https://github.com/terraytherapeutics/terramax.git
  license: Apache-2.0
  official: true
  kind: research-library
  description: >-
    TerraMax — pretrained joint predictions for scalable batch Bayesian
    optimization of molecular designs; companion code to arXiv:2511.10590.
- language: python
  registry: github
  name: streamlit-ketcher
  url: https://github.com/terraytherapeutics/streamlit-ketcher
  install: pip install git+https://github.com/terraytherapeutics/streamlit-ketcher.git
  license: Apache-2.0
  official: false
  kind: fork
  description: >-
    Terray fork of the streamlit-ketcher molecule-editor component with local
    tweaks; the upstream package is published to PyPI by its original authors.
x-evidence:
  fetched: '2026-08-02'
  sources:
  - url: https://api.github.com/orgs/terraytherapeutics/repos
    http_status: 200
  - url: https://pypi.org/pypi/coati/json
    http_status: 200
    note: unrelated project (Intelimetrica PowerPoint tool)
  - url: https://pypi.org/pypi/terramax/json
    http_status: 404
  - url: https://registry.npmjs.org/terray
    http_status: 404