OmniBioAI Studio

Desktop orchestration platform for AI-powered bioinformatics computation

OmniBioAI Studio is an Electron desktop app that launches and manages the full OmniBioAI stack — locally, on HPC clusters, or in the cloud — with a single click.

System documentation

System architecture — boundaries, flows, deployment topology, and the complete 40-service catalog
Security hardening — production Compose security and development overrides
Integration tests — end-to-end test setup
Operations scripts — backup, validation, and deployment utilities

✨ What's New in v0.7.0

🔐 Unified license key system (OMNI-XXXX-XXXX-XXXX-XXXX) — one key works for web + desktop, auto-creates user on first validation, same JWT as OAuth login
🌐 webstudio.omnibioai.org fully working — license key login enforced, all workbench pages loading correctly, service worker fixed, nginx routes fixed (/license/, /roles/), Control Center JWT cookie fallback
👥 Team expansion — Dr. Rajnish Kumar (Scientific Consultant, SR University), Praveen C.V. Raghavulu (Scientific Advisor, KUMC); About page restructured as company team page
🛠 12,110 bioinformatics tools (up from 11,577) — 100+ new HTTP API tools added, all tools validated (100% clean), 0 duplicates, 0 ToolSpec errors
🤖 Tool selection AI improved — GPU enabled for Ollama (57x faster: 120s → 2.1s), accuracy 0% → 60%, Recall@K 60% → 85%
📦 1,000 ARM64 SIF images
🔧 Control Center web service added
💰 Billing service integrated
🛡️ Admin Console live at admin.omnibioai.org

v0.6.0-beta ✅

🌐 Web version at webstudio.omnibioai.org — no installation required
🔐 SSO & OAuth2 — Google, GitHub, and Microsoft sign-in
🛡️ Cloudflare Access — secure invite-only email whitelist
🧬 12,000+ bioinformatics tools across HTTP APIs, ARM64, x86, and Kubernetes
📦 1,120+ container images (320 Docker + 800 ARM64 SIF) hosted on GHCR and Hugging Face
🔌 225+ bioinformatics & AI plugins covering scRNA-seq, WGS, WES, proteomics, spatial, and more
⚙️ 600+ workflow bundles for Nextflow, WDL, CWL, and Snakemake
🤖 36M PubMed abstracts indexed with a 150-domain RAG pipeline (PubMedBERT, FAISS, BM25, RRF, Neo4j)
🤗 One-click Hugging Face Push from the Model Registry
📊 Live platform metrics dashboard with architecture, service health, and coverage

v0.4.0-beta ✅

Version unification — all UI components, sidebar, badge, logs, and settings now consistently report v0.4.0-beta
23 services fully operational — all layers (Data, Security Control Plane, Execution, AI, Developer) green
1,010 registered tools — confirmed live in Jobs → Registered Tools panel (platform-wide tooling, including HPC/cloud/orchestration integrations, totals 11,000+ — see Bioinformatics Tools section below)
7 execution servers — local_real, slurm_local, aws_batch_prod, aws_batch_demo, azure_batch_demo, gcp_batch_demo, enrichment_remote
claude-sonnet-4-20250514 as default orchestrator model in LLM configuration
IDE Services all RUNNING — JupyterLab (:8888), RStudio (:8787), VS Code Server (:8883)
Beta Cloud mode — connects to webstudio.omnibioai.org; MySQL, Workbench, TES, Ollama tunnels all reachable
Report Bug modal — title, description, email, severity (Low / Medium / High / Critical) with Submit Bug Report

v0.3.0-beta ✅

IDE Services — JupyterLab, RStudio, and VS Code Server managed directly from Studio UI
IDE Layer — dedicated section on Services page with per-container lifecycle management
Launcher backend — Express API using Docker socket for IDE container control; ARM64-compatible
Unified Grafana metrics dashboard embedded in Studio
Full observability stack: cAdvisor + redis-exporter + django-prometheus
OmniBioAI dark theme on Grafana and Prometheus
Auto-generated secrets on first launch via crypto.randomBytes
Grafana service account token auth (anonymous access disabled)
Zero npm vulnerabilities (Electron 28→41, vite 5→8, all CVEs resolved)
DMG + AppImage + EXE installers via GitHub Actions
Public beta announcement + Cloudflare-integrated beta signup

v0.2.0-beta ✅

License key system (OMNI-XXXX-XXXX-XXXX-XXXX, 30-day trial)
Sentry error tracking + in-app bug report button
Cython IP protection (.so compiled binaries)
MySQL-backed license server
1010+ bioinformatics tools (510 HTTP API + 500 Slurm)
Windows NSIS .exe installer
Zero-trust JWT authentication, RBAC/ABAC policy engine
HPC quota governance + async audit logging via Redis Streams

v0.1.0-beta ✅

Full local stack launch with containerized services
Live service health monitoring
Docker image dashboard
Dev Hub with knowledge graph + RAG UI
Mode-aware startup: Local / HPC / Cloud / Hybrid
LLM configuration: Ollama + Claude API + OpenAI
Cloud execution: AWS Batch / Azure Batch / GCP Batch / Kubernetes
HPC execution: Slurm / PBS / LSF via TES

🖥 Screenshots

Runtime Mode — Service Health

Beta Cloud selected — MySQL, Redis, TES, Workbench UP; Ollama initializing

Workbench — Module Overview

Quick access to all 44 modules across 6 sections

LLM Configuration

Local Ollama (deepseek-coder), Claude API, OpenAI/Codex, and runtime orchestration settings

Cloud Configuration

AWS, Azure, GCP, and Kubernetes execution backends with full credential management

HPC Configuration

Slurm scheduler, SSH connection, GPU jobs, TES remote execution, filesystem & runtime settings

Launch — Execution Console

Connected to Beta Cloud — all tunnels reachable, runtime summary visible

Services — Full Stack

40 Compose services across Data, Security Control Plane, Execution, AI, and Developer layers

IDE Services

JupyterLab, RStudio, VS Code Server — all RUNNING, managed via Launcher :5190

Live Logs

Real-time log stream — 7 entries, filterable by service, live streaming

Jobs — TES Execution Engine

1,010 registered tools · 7 execution servers (local, Slurm, AWS, Azure, GCP, enrichment_remote)

Settings

Data directories, service ports, Docker compose file, security, About panel

Bug Report

In-app bug reporting with title, description, email, and severity selector

📦 Downloads

Platform	File	Requirements
macOS Apple Silicon (M1/M2/M3/M4)	`OmniBioAI-Studio-arm64.dmg`	macOS 12+
macOS Intel	`OmniBioAI-Studio-x64.dmg`	macOS 12+
Linux x86_64 AppImage	`OmniBioAI-Studio.AppImage`	Ubuntu 20.04+
Linux x86_64 DEB	`OmniBioAI-Studio.deb`	Ubuntu / Debian
Linux x86_64 RPM	`OmniBioAI-Studio.rpm`	RHEL / Fedora
Linux ARM64 AppImage	`OmniBioAI-Studio-arm64.AppImage`	aarch64, Ubuntu 20.04+
Linux ARM64 DEB	`OmniBioAI-Studio-arm64.deb`	Ubuntu / Debian ARM64
Linux ARM64 RPM	`OmniBioAI-Studio-arm64.rpm`	RHEL / Fedora ARM64
Windows	`OmniBioAI-Studio-Setup.exe`	Windows 10/11 + WSL2

Download from: https://github.com/OmniBioAI/omnibioai-studio/releases/latest

📊 Live Platform Proof

Real-time architecture, codebase metrics, coverage, and service health are publicly viewable at:

control.omnibioai.org

🔐 Security Control Plane

All requests are enforced through a zero-trust pipeline:

Internet / Client
↓
api-gateway :8080 ← single entry point, JWT enforcement
↓
auth-service :8001 ← JWT validation + Redis cache (TTL=300s)
↓
policy-engine :8002 ← RBAC/ABAC authorization decision
↓
hpc-policy-engine :8003 ← GPU/CPU quota check (compute requests only)
↓
target service (workbench / tes / toolserver / rag)
↓
security-audit :8004 ← async audit log → Redis Streams (never blocks)

Layer	On failure
Auth	FAIL CLOSED → HTTP 401
Policy	FAIL CLOSED → HTTP 403
HPC quota	FAIL CLOSED → HTTP 403
Audit	FAIL OPEN → ignored

🔐 Browser Authentication

The diagram above is the server-side request pipeline; this section covers how the Studio SPA itself holds and presents a session in the browser. Login/refresh/logout all go through the standard omnibioai-auth endpoints — see omnibioai-auth's README for the full token model.

Session cookies

Studio's web build still manages its own session client-side — localStorage["omnibioai_access_token"] and localStorage["omnibioai_refresh_token"] — rather than relying on omnibioai-auth's server-set, HttpOnlyomnibioai_session cookie (see that repo's Session Cookies section). It additionally mirrors the access token into a non-HttpOnly, JS-writable cookie of the same name (omnibioai_access_token, SameSite=Lax, Secure over HTTPS) purely so an embedded iframe can authenticate — see iframe authentication below. This is a distinct mechanism from omnibioai-auth's own session cookie: same-looking pattern, different cookie, different owner (browser JS here, vs. server-set and HttpOnly there).

Control Center integration

omnibioai-control-center's Admin tab, when embedded under Studio's own origin, reads this same localStorage["omnibioai_access_token"] key — an existing Studio login is recognized automatically with no separate sign-in, since both apps share one browser origin in that deployment path. See Control Center's Authentication section for the admin-side detail.

iframe authentication

Control Center is embedded via <iframe src="/_svc/control"> in the web build (an Electron <webview> in the desktop build). An iframe's initial document navigation can't carry a custom Authorization header, so docker/nginx-router.conf falls back to the mirrored omnibioai_access_token cookie: it maps the cookie's value into a synthesized Authorization: Bearer <token> header for the /_svc/control location and for the shared /internal/auth/verify subrequest, so the iframe's first request authenticates even though no JavaScript ran inside it yet. Subsequent same-origin fetch/XHR calls made from inside the iframe read localStorage directly, same as the parent page. (Grafana is also embedded via iframe/webview, but authenticates with its own session cookie, unrelated to this mechanism.)

Refresh flow

refresh() posts to /auth/refresh with the refresh_token stored in localStorage. The root App component calls it once on startup and then every five minutes while a refresh token is present, keeping the shared access token fresh for the Studio UI and embedded services. Logout and a failed refresh clear the stored session; individual API clients may still handle a 401 according to their own behavior. This remains a client-side localStorage flow and is distinct from the cookie-based session used by omnibioai-control-center.

Logout

logout() posts both the stored refresh_token and access_token to /auth/logout (fails open on a network error), then always clears both localStorage keys and the mirrored omnibioai_access_token cookie — regardless of whether the server call itself succeeded.

🖥 Services

Data Layer

Service	Port	Image
MySQL	:3306 (internal only in production/release — see below)	mysql:8.0
Redis	:6379 (internal only in production/release — see below)	redis:7-alpine

Production/release (docker-compose.release.yml, the config packaged into the Electron app): MySQL and Redis are not published to the host — reachable only inside the Compose network, as mysql:3306 / redis:6379. Every other service still addresses them exactly that way.

Development: the local dev stack (docker-compose.yml) still publishes both directly (:3306 / :6380) for convenience, as it always has. To get the same local access against the release stack instead, layer the explicit docker-compose.release.dev-ports.yml overlay:

docker compose -f docker-compose.release.yml -f docker-compose.release.dev-ports.yml up -d

This overlay binds to 127.0.0.1 only, not 0.0.0.0, and is never bundled into the packaged app or referenced by its startup path — it has to be opted into explicitly. See SECURITY-COMPOSE-HARDENING.md for the full rationale.

Security Control Plane

Service	Port	Image
API Gateway	:8080	ghcr.io/omnibioai/omnibioai-api-gateway:latest
Auth Service	:8001	ghcr.io/omnibioai/omnibioai-auth:latest
Policy Engine	:8002	ghcr.io/omnibioai/omnibioai-policy-engine:latest
HPC Policy Engine	:8003	ghcr.io/omnibioai/omnibioai-hpc-policy-engine:latest
Security Audit	:8004	ghcr.io/omnibioai/omnibioai-security-audit:latest

Execution Layer

Service	Port	Image
Workbench	:8000	ghcr.io/omnibioai/omnibioai-app:latest
TES	:8081	omnibioai-tes-local
ToolServer	:9090	ghcr.io/omnibioai/omnibioai-toolserver:latest
Model Registry	:8095	ghcr.io/omnibioai/omnibioai-model-registry:latest
LIMS	:7000	ghcr.io/omnibioai/omnibioai-lims:latest
Control Center	:7070 (localhost-only, JWT-gated via nginx `/_svc/control`)	ghcr.io/omnibioai/omnibioai-control-center:latest
Control Center Web	127.0.0.1:5174 (frontend dev target, built from `omnibioai-control-center`'s Dockerfile)	build-only
Billing Service	:8005	build: `../omnibioai-billing` (`Dockerfile`)
Billing Worker	— (background consumer, no exposed port)	build: `../omnibioai-billing` (`Dockerfile.worker`)
Workflow Bundles	:8098	ghcr.io/omnibioai/omnibioai-workflow-bundles:latest
Tool Images	:8097	ghcr.io/omnibioai/omnibioai-tool-images:latest

AI Layer

Service	Port	Image
Ollama	:11434	ollama/ollama
RAG	:8090 (external) / :8096 (internal)	ghcr.io/omnibioai/omnibioai-rag:latest
Dev Hub	:8082	ghcr.io/omnibioai/omnibioai-dev-hub:latest
Neo4j	:7474 / :7687	neo4j:5.15

Developer Layer

Service	Port	Image
Launcher	:5190	ghcr.io/omnibioai/omnibioai-launcher:latest

IDE Services (managed via Launcher :5190)

Service	Port	Stack
JupyterLab	:8888	Full bioinformatics stack (scanpy, DESeq2, scVelo, cellxgene…)
RStudio Server	:8787	R + Bioconductor (Seurat, DESeq2, scran, monocle3, tidyverse)
VS Code Server	:8883	Python + R + Nextflow + WDL extensions

Observability & Platform Infrastructure

Service	Port	Image
Grafana	:3000	grafana/grafana:latest
Prometheus	internal only, via `/_svc/prometheus`	prom/prometheus:latest
cAdvisor	:8585	gcr.io/cadvisor/cadvisor:latest
Redis Exporter	:9121	oliver006/redis_exporter:latest
Node Exporter	host network, no published port	prom/node-exporter:latest
License Server	:8099	internal build
OPA (Open Policy Agent)	:8181	openpolicyagent/opa:latest
Videos	:8086	ghcr.io/omnibioai/omnibioai-videos:latest
Web UI	served as static files behind Nginx Router	build: `Dockerfile.web` (this repo)
Nginx Router	:80	nginx:latest

🧰 Bioinformatics Tools (11,000+)

1,010 tools were recorded in the latest deployment snapshot (verify the current value in Jobs → Registered Tools). The broader platform catalog — including execution, cloud, HPC, and orchestration tooling across every service — is reported as 11,000+.

Execution Servers (7)

Server ID	Adapter
local_real	local
slurm_local	slurm
aws_batch_prod	aws_batch
aws_batch_demo	aws_batch
azure_batch_demo	azure_batch
gcp_batch_demo	gcp_batch
enrichment_remote	http_toolserver

HTTP API Tools (510)

Direct REST integrations — no compute needed: Ensembl, NCBI, ClinVar, gnomAD, UniProt, AlphaFold, KEGG, Reactome, PubMed, ChEMBL, DrugBank, CellxGene, HMDB, and 280+ more.

Slurm/HPC Tools (500)

Compute tools: BWA, STAR, HISAT2, GATK, DeepVariant, DESeq2, Seurat, Scanpy, PyTorch, MSFragger, and 90+ more.

🧬 Workbench Modules (44 total)

Platform Services (14 modules)

Getting Started · Video Tutorials · Workbench Dashboard · Control Center · LIMS · Model Registry · RAG/Lit AI · TES/Jobs · Tool Images · Launcher · Workflows · Dev Hub · Metrics · Grafana

Security Control Plane (6 modules)

API Gateway · Auth Service · Policy Engine · HPC Policy · Security Audit · OPA

Core Platform (6 modules)

Home · OnboardAI · OmniBioAgent · Job Monitor · Plugin Manager · Admin

Workflows (6 modules)

Workflow Runner · Workflow Builder · Agent Studio · Pipeline Dashboard · Multi-Agent Orchestrator · Workflow Compiler

Omics Analysis (6 modules)

RNA-Seq · Single Cell (scRNA-Seq) · Exome Analysis · FASTQ QC · Proteomics · Metabolomics

AI & Intelligence (6 modules)

Drug Target AI · Literature AI · Pathway Enrichment · Bio Hypothesis · Literature Summarizer · Bio Narrator AI

🖥 Runtime Modes

Mode	Status	Description
Beta Cloud	✅ Available	Connects to `webstudio.omnibioai.org` — no local Docker needed
Local	✅ Available	Docker-based local stack; GPU/CPU support depends on host configuration
HPC	Coming soon	Slurm / PBS / LSF, Apptainer remote execution
Cloud	Coming soon	AWS Batch / Azure Batch, elastic auto-scaling
Hybrid	Coming soon	Multi-backend orchestration, policy-driven scheduling

🤖 LLM Configuration

Provider	Model	Notes
Ollama (local)	deepseek-coder:latest	Default local model; GPU-accelerated
Embedding	nomic-embed-text	Local embedding model
Claude API	claude-sonnet-4-20250514	Default orchestrator model
OpenAI / Codex	gpt-4o	Optional cloud fallback

Runtime options: Offline-first mode, Enable RAG (vector retrieval), Default Orchestrator Model selector.

☁️ Cloud Configuration

Provider	Features
Amazon Web Services	AWS Batch, IAM access keys, S3, region selector
Microsoft Azure	Azure Batch, subscription ID, tenant ID, Blob storage
Google Cloud Platform	GCP Batch, project ID, service account JSON, Cloud Storage
Kubernetes	kubeconfig path, context, namespace, SIF base URL, job prefix
Future: Databricks Workflows	—
Future: Slurm Cloud Bridge	—

🏗 HPC Configuration

Setting	Value
Scheduler	Slurm (dropdown: Slurm / PBS / LSF)
GPU Jobs	CUDA cluster support
Remote Execution	via TES protocol
SSH Hostname	hpc.university.edu (configurable)
SSH Port	22
Private Key	~/.ssh/id_rsa
Shared Mount	/shared/projects
Apptainer Path	/usr/bin/apptainer
Default Partition	gpu

📊 Observability

Grafana Dashboards (4)

OmniBioAI Services — health, request rate, latency, container resources
OmniBioAI Platform Overview — full platform architecture metrics
OmniBioAI LIMS — lab information management metrics
OmniBioAI RAG — query latency and throughput

Prometheus Scrape Targets (7)

workbench:8000 · lims:7000 · rag:8096 · auth-service:8001 · control-center:7070 · cadvisor:8080 · redis-exporter:9121

📋 System Requirements

Component	Minimum	Recommended
RAM	16 GB	32 GB (64 GB with local LLM)
Disk	50 GB free	100 GB free
Docker	Engine 24+	Docker Desktop
OS	Ubuntu 20.04+, macOS 12+, Windows 10/11 (WSL2)	Ubuntu 22.04+
GPU	Optional	NVIDIA + nvidia-container-toolkit

Also required: jq (sudo apt install jq), Docker Compose v2 (included with Engine 24+)

🚀 Quick Start

Beta Cloud (no Docker needed)

Download installer for your platform from Releases
Launch OmniBioAI Studio and enter your license key
Select Beta Cloud on the Mode page
Click through the setup wizard (Steps 1–5)
Click Launch — tunnels connect to webstudio.omnibioai.org automatically

Local Stack (Docker)

git clone https://github.com/OmniBioAI/omnibioai-studio
cd omnibioai-studio
cp .env.example .env
# Edit .env — set DATA_DIR, WORK_DIR, and secrets
docker compose up -d

From Source

npm install
npm run dev # development mode (Vite + Electron)
npm run build # AppImage (Linux)
npm run build:mac # DMG (macOS)
npm run build:win # EXE (Windows)

⚙️ Settings

Data Directories

Path	Purpose
Data Directory	PubMed abstracts, FAISS indexes, RAG data
Work Directory	Workflow results, runs, outputs

Expected layout:

data/PubMed/Index/<study>/pubmed_index.faiss
work/workflow_runner.runs/, work/uploads/, work/objects/

Service Ports (configurable)

Service	Default Port
Workbench	8000
TES	8081
ToolServer	9090

Changing ports requires a full stack restart.

Docker

Compose file: docker-compose.yml (repo root — this is what docker compose up -d under "Quick Start" runs; docker/docker-compose.yml is a separate, much smaller stub, not the one the full stack uses)
Data Dir mounted as /data in all containers
Work Dir mounted as /workspace/work in all containers

About (v0.7.0)

Field	Value
Studio Version	v0.7.0
Electron	Electron + Vite
Node.js	See `package.json` toolchain
Platform	Runtime-dependent (Linux, macOS, or Windows)
Status	Beta

🔑 License

OmniBioAI Studio requires a license key for first launch.

Format:OMNI-XXXX-XXXX-XXXX-XXXX (30-day trial)
Get access:omnibioai.org/#request
Offline grace period: 7 days after initial validation
Beta users receive a GitHub token automatically with their license key

🔑 Environment Variables

Copy .env.example to .env and fill in values — this is the actual, current set (cp .env.example .env):

# ── Network ────────────────────────────────────────────
HOST_IP=0.0.0.0
# ── Paths (absolute paths on host) ─────────────────────
MACHINE_DIR=/path/to/your/machine/dir
WORKSPACE_HOST=/path/to/omnibioai
WORK_DIR=/path/to/omnibioai
DATA_DIR=/path/to/data
DB_INIT_DIR=/path/to/db-init
VIDEO_DIR=/path/to/omnibioai-videos/content
# ── Database ───────────────────────────────────────────
MYSQL_ROOT_PASSWORD=change-me-in-production # auto-generated on first launch
MYSQL_DEFAULT_DB=omnibioai # optional; defaults to omnibioai in compose# ── Auth ───────────────────────────────────────────────
AUTH_SECRET_KEY=change-me-in-production # auto-generated on first launch
LICENSE_SECRET=change-me-in-production # auto-generated on first launch# ── LIMS ───────────────────────────────────────────────
LIMS_USERNAME=admin
LIMS_PASSWORD=change-me
LIMS_REFRESH_TOKEN=
# ── LLM / AI (optional) ────────────────────────────────
ANTHROPIC_API_KEY=
OPENAI_API_KEY=
RAGBIO_API_KEY=
# ── Monitoring (optional) ──────────────────────────────
SENTRY_DSN= # empty disables in-app bug reporting
SENTRY_ENVIRONMENT=production
SENTRY_RELEASE=1.0.0
DISCORD_WEBHOOK_URL=
DISCORD_ALERT_WEBHOOK_URL= # only fires for new high-severity known-issue entries# ── GitHub (for pulling private images) ────────────────
GHCR_PULL_TOKEN=
GF_ADMIN_PASSWORD=omnibioai # auto-generated on first launch# ── IDE Services ───────────────────────────────────────
JUPYTER_TOKEN=omnibioai
RSTUDIO_PASSWORD=omnibioai
VSCODE_PASSWORD=omnibioai

OMNIBIOAI_DEV_MODE is a separate, CI-only flag (.github/workflows/ci.yml) — it is not a .env/compose setting and doesn't affect a local docker compose up.

🐛 Bug Reporting

Click the 🐛 Report Bug button in the Studio UI at any time.

Fields: Bug title · Description · Email (optional) · Severity (Low / Medium — Affects workflow / High / Critical)

Reports are sent to our dashboard. Response within 24 hours during beta. Disable with SENTRY_DSN= (empty) in .env.

🔗 OmniBioAI Ecosystem

Repository	Role
`omnibioai`	Main Django workbench + 351 plugins
`omnibioai-api-gateway`	Zero-trust API gateway
`omnibioai-auth`	JWT authentication service
`omnibioai-policy-engine`	RBAC/ABAC authorization
`omnibioai-hpc-policy-engine`	GPU/CPU quota governance
`omnibioai-security-audit`	Async audit logging
`omnibioai-tes`	Task Execution Service
`omnibioai-toolserver`	FastAPI tool API
`omnibioai-lims`	Lab data management
`omnibioai-model-registry`	ML model versioning
`omnibioai-control-center`	Health + image dashboard
`omnibioai-rag`	PubMed RAG pipeline
`omnibioai-dev-hub`	Knowledge graph + embeddings
`omnibioai-workflow-bundles`	WDL/Nextflow/Snakemake bundles
`omnibioai-launcher`	SDK UI + IDE container lifecycle API
`omnibioai_sdk`	Python SDK client
`omnibioai-security-sdk`	Security SDK for service auth
`omnibioai-design-tokens`	Shared design tokens and theme
`omnibioai-ui`	Shared UI component library
`omnibioai-landing`	Public-facing landing page

🗺 Roadmap

Version	Status	Highlights
v0.1.0-beta	✅ Released	Local stack, health monitoring, Dev Hub, LLM configuration
v0.2.0-beta	✅ Released	License system, zero-trust security, 1,010 tools, Windows installer
v0.3.0-beta	✅ Released	IDE Services, Grafana observability, auto-secrets, npm security
v0.4.0-beta	✅ Released	Version unification, 23 services, 7 execution servers, Claude Sonnet 4
v0.5.0-beta	✅ Released	225+ plugins, 36M-abstract RAG index, 1,120+ container images, full beta launch
v0.6.0-beta	✅ Released	Web version (`webstudio.omnibioai.org`), SSO/OAuth2 (Google/GitHub/Microsoft), Cloudflare Access, 800 ARM64 SIF images, 12,000+ tools, Hugging Face integration, Model Registry HF push button
v0.7.0	✅ Current	Unified License Key (one `OMNI-XXXX` key for web + desktop, auto-creates user, same JWT as OAuth login), `webstudio.omnibioai.org` fully working end-to-end, 12,110 tools (100+ new HTTP API tools), Tool Selection AI 57x faster with GPU-enabled Ollama (accuracy 0%→60%, Recall@K 60%→85%), 1,000 ARM64 SIF images, Control Center web service, Billing service integrated, Admin Console at `admin.omnibioai.org`, team expansion (Dr. Rajnish Kumar, Praveen C.V. Raghavulu)
v0.8.0	🔜 Planned	Enterprise Platform — expand the ML portfolio to 20 trained ML models on Hugging Face, HIPAA compliance reporting, SAML/enterprise SSO, multi-user workspaces, team management, usage analytics, mobile-responsive UI, and OmniBioAgent v2 (tool-aware responses, agentic workflow suggestions, multi-step reasoning, context-aware analysis)
v0.9.0	🔜 Planned	Enterprise Admin Console expansion — advanced organization administration, enterprise governance, deeper billing and subscription management, advanced usage analytics, administrative workflows, platform operations, security administration, and enterprise-scale management capabilities

🐛 Known Issues (Beta)

System MySQL/Redis must be stopped before starting: sudo systemctl stop mysql redis-server
GITHUB_TOKEN must be set manually for private image pull
macOS DMG not yet code-signed (GateKeeper warning expected)
Windows installer not yet code-signed
First launch requires internet for license validation; 7-day offline grace period after
cAdvisor requires privileged mode and /dev/kmsg device access
Prometheus not exposed directly — access only via /_svc/prometheus
Control Center (/_svc/control) requires valid JWT; port 7070 bound to localhost only
Billing service backend (billing-service, :8005) is deployed and DB-backed, but has no served production frontend — control-center-web (the billing/subscriptions/entitlements UI) is not wired into any docker-compose file or nginx route in this deployment yet
Two license-validation backends currently coexist: the legacy standalone license_server.py (license-server, :8099, its own MySQL DB) and the unified /license/validate endpoint on omnibioai-auth (:8001), which is what web and desktop actually call for the OMNI-XXXX login flow. The legacy server is still built and deployed but appears superseded — pending a decision on formal decommission

🛠 Maintenance Scripts

Script	Description	Schedule
`scripts/backup-mysql.sh`	Dumps all DBs to compressed `.sql.gz`, 7-day rotation	Daily at 4am
`scripts/check-env.sh`	Validates `.env` secrets before stack start	Before `docker compose up`
`omnibioai-control-center/scripts/run_coverage_host.py`	Rebuilds ecosystem coverage report	Daily at 2am
`omnibioai-dev-hub/scripts/check_and_reindex.sh`	Rebuilds RAG FAISS index on new Studio release	Hourly (checks for new release tag)

04***/home/manish/Desktop/machine/omnibioai-studio/scripts/backup-mysql.sh >> /home/manish/Desktop/machine/work/backups/omnibioai-backup.log 2>&102***python3 /home/manish/Desktop/machine/omnibioai-control-center/scripts/run_coverage_host.py --root /home/manish/Desktop/machine >> /home/manish/Desktop/machine/work/backups/omnibioai-coverage.log 2>&10****/home/manish/Desktop/machine/omnibioai-dev-hub/scripts/check_and_reindex.sh >> /home/manish/Desktop/machine/work/backups/omnibioai-reindex.log 2>&1

📄 License

Apache 2.0 — see LICENSE

👤 Author

Manish Kumar — GitHub · omnibioai.org

OmniBioAI Studio is not a bioinformatics tool — it is a desktop orchestration system for distributed, secure, AI-native scientific computation.

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
backend		backend
build		build
db-init		db-init
docker		docker
docs		docs
electron		electron
monitoring		monitoring
packages		packages
public		public
scripts		scripts
src		src
tests		tests
.dockerignore		.dockerignore
.env.example		.env.example
.env.web		.env.web
.gitignore		.gitignore
CHANGELOG.md		CHANGELOG.md
CONTRIBUTING.md		CONTRIBUTING.md
DEPLOYMENT.md		DEPLOYMENT.md
Dockerfile.web		Dockerfile.web
LICENSE		LICENSE
README.md		README.md
SECURITY-COMPOSE-HARDENING.md		SECURITY-COMPOSE-HARDENING.md
docker-compose		docker-compose
docker-compose-release.yml		docker-compose-release.yml
docker-compose.release.dev-ports.yml		docker-compose.release.dev-ports.yml
docker-compose.release.yml		docker-compose.release.yml
docker-compose.yml		docker-compose.yml
electron-builder.json		electron-builder.json
index.html		index.html
new_packages.txt		new_packages.txt
old_packages.txt		old_packages.txt
package-lock.json		package-lock.json
package.json		package.json
postcss.config.js		postcss.config.js
sif_skipped.txt		sif_skipped.txt
tailwind.config.js		tailwind.config.js
test_integration.sh		test_integration.sh
vite.config.js		vite.config.js

Uh oh!

Latest commit

History

Folders and files

Repository files navigation