OmniBioAI v1.2 Documentation

OmniBioAI brings project context, analysis execution, workflows, tools, and result handling into one platform. This site explains how to use it, how it is built and extended, how it is operated, and — just as importantly — what is and is not established about its quality, security, and reproducibility.

Where to start

I want to…	Start with
Get oriented	Getting Started and Platform Architecture
Use the platform: sign in, run an analysis or workflow, read results	User Guide, beginning with Getting Started
Understand how the platform is put together	Architecture
Find what is available: plugins, integrations, tools, workflows, reference data, biomedical knowledge	Catalogs
Build against or extend OmniBioAI	Developer Guide — start with the Repository Map
Call the APIs	API Reference and the API Guide
Deploy, configure, or operate a deployment	Platform Administration
Diagnose something that is not working	Troubleshooting
Contribute code or documentation	Contributing and Source Freshness
Check what is tested, secured, reproducible, and monitored	Quality & Trust

Catalogs at a glance

These figures are read from the generated catalogs, not typed into this page. Each one names its unit and the evidence level its source supports; see Catalogs for what the numbers mean.

• Workbench Plugin Catalog

  501 registered plugins

  Evidence level CONFIGURED

  Registered in source. Not tested, deployed or operationally verified.

• Integrations

  66 catalogued integrations (a subset of the plugins)

  Evidence level CONFIGURED

  28 declare write or high-impact actions. Enabled is not operationally verified.

• Reference Data: database plugins

  129 reference-database plugins

  Evidence level CONFIGURED

  125 IMPLEMENTED, 2 TEMPLATE_SCAFFOLD, 2 UNKNOWN in source. Not a count of supported databases.

• Reference Data: organism assets

  12 organism directories (local snapshot, separate from the plugins)

  Evidence level MEASURED_ON_DISK

  Directory listing at a stated time; says nothing about completeness or provenance.

• TES Tool Catalog

  12,276 configured tool definitions

  Evidence level CONFIGURED

  9,061 carry the source tag unverified_command. Catalogued is not executed or verified.

• TES Execution Backends

  8 configured backend definitions

  Evidence level CONFIGURED

  Definitions in source configuration, not deployed or reachable backends.

• Workflow Catalog

  869 registered workflows (of 879 candidate directories)

  Evidence level CONFIGURED

  A valid manifest and README, not an executed or verified workflow. Maintainer-reported (not verified from repository evidence): 1000+ workflows exist in total; this count is the ones registered here.

• API Reference

  251 source-derived routes in 8 services (first slice)

  Evidence level IMPLEMENTED_IN_SOURCE

  110 are itemised publicly. A lower bound found in source, not a platform endpoint count.

• Biomedical Knowledge

  282 active domain indexes

  Evidence level MEASURED_ON_DISK

  Mixed 768/1024 dimension; 1024 is the configured target, not the current state.

How to read the labels

A catalog entry says what is registered or configured in source. It does not say the entry has been tested, deployed, or operationally verified. Evidence levels are defined in Quality & Trust and on the operator guidance page.