dbt-Workbench: Open Source dbt UI

dbt-Workbench is an open source dbt UI for lineage visualization, run orchestration, catalog and docs workflows, and environment management without vendor lock-in. It is designed for local, on-prem, and air-gapped deployments where teams need full control.

Keywords: dbt UI, open source dbt, dbt lineage, data lineage tool, dbt catalog, dbt scheduler, self-hosted dbt, air-gapped dbt

Documentation: https://rezer-bleede.github.io/dbt-Workbench/

---

✨ Highlights

• Unified dbt control plane for artifacts, runs, and environments
• Interactive lineage at model and column granularity with a deterministic D3/dagre layout, pan/zoom, and grouping controls
• Workspace-aware multi-project setup with strict path scoping
• Secure by design with optional JWT authentication + RBAC
• Extensible plugin system and Git-integrated workspace

---

🧭 Table of Contents

• Screenshots
• Quickstart
• Run with Docker Compose
• Multi-Project Workspaces
• Local Development
• Project Structure
• Authentication & RBAC
• Environment Variables
• Features Overview
• API Reference
• Testing
• Documentation

---

📸 Screenshots

---

🚀 Quickstart

Prerequisites

• Docker
• Docker Compose

---

🐳 Run with Docker Compose

docker-compose up --build
# or
docker compose up --build

Demo dbt Project

The repository includes a ready-to-run demo project in ./dbt_project with raw, staging, and mart models so dbt run succeeds immediately against DuckDB. The default workspace that the backend bootstraps uses the same model layout, giving you consistent sample data for lineage, runs, and catalog exploration.

Services

• UI: http://localhost:3000
• API: http://localhost:8000
• API Docs: http://localhost:8000/docs (Swagger UI)

Mounting dbt Artifacts

The backend mounts:

./sample_artifacts → /app/dbt_artifacts

Replace sample_artifacts with your dbt target/ directory containing:

• manifest.json
• run_results.json
• catalog.json

For the in-app Docs page, include the full contents from dbt docs generate (the index.html and accompanying assets) inside your mounted target/ directory. The viewer will automatically serve that site from the latest artifacts so you can browse the complete dbt documentation without leaving the UI.

The UI will load and display real metadata from your dbt project automatically.

---

🧭 Multi-Project Workspaces

dbt-Workbench can manage multiple isolated projects in a single instance. Each workspace gets its own repository folder under the configured GIT_REPOS_BASE_PATH, independent artifacts storage, and per-workspace settings. File operations are hard-scoped to the active workspace root to prevent cross-project access or path traversal, and switching workspaces refreshes the active project context across the UI and API.

For local development, set a dedicated base path for repositories:

export GIT_REPOS_BASE_PATH=$(pwd)/data/repos

Each workspace will use a subdirectory under that path (e.g., data/repos/<workspace-key>), keeping source files, artifacts, and run history isolated by project.

When authentication is disabled you can still switch the active project by sending the X-Workspace-Id header on any API call. The frontend persists the last-selected workspace locally so reloading the UI keeps the correct project context without leaking data between workspaces.

dbt-Workbench automatically bootstraps a local Demo Project with its own git repository on first launch so you can explore the UI immediately. The Projects & Version Control page lets you add local-only projects (each with their own repo) or connect remotes; on subsequent visits the UI loads whichever project you last activated.

---

🔧 Local Development

Backend (FastAPI)

cd backend
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

Database Schema Compatibility

On startup, the backend runs lightweight schema checks to keep legacy databases compatible with new features. For example, it will add the runs.logs column if it is missing so existing installations can keep serving run history without a manual migration step.

Frontend (React + TypeScript + Vite)

cd frontend
npm install
npm run dev -- --host --port 3000

Set the API base URL if needed:

VITE_API_BASE_URL = http://localhost:8000

---

📁 Project Structure

dbt-Workbench/
│
├── backend/               # FastAPI service for metadata + execution engine
│   ├── app/
│   │   ├── api/routes/    # API endpoint handlers
│   │   ├── core/          # Config, auth, plugins
│   │   ├── database/      # SQLAlchemy models and services
│   │   ├── schemas/       # Pydantic request/response models
│   │   ├── services/      # Business logic services
│   │   └── main.py        # FastAPI application entry
│   └── requirements.txt
├── frontend/              # React + TS + Vite UI
│   ├── src/
│   │   ├── components/    # Reusable UI components
│   │   ├── context/       # React contexts (Auth, etc.)
│   │   ├── pages/         # Page components
│   │   └── services/      # API service clients
│   └── package.json
├── plugins/               # Plugin directory (manifest + backend/frontend assets)
├── sample_artifacts/      # Demo dbt artifacts
├── docker-compose.yml     # Full stack orchestration
├── ARCHITECTURE.md        # System architecture documentation
├── PLUGIN_SYSTEM.md       # Plugin system specification
├── CONTRIBUTING.md        # Contribution guidelines
├── ROADMAP.md             # Development roadmap
└── README.md

---

🔐 Authentication & RBAC

Authentication Modes

Setting	Behavior
AUTH_ENABLED=false (default)	No login required, all users have Admin access
AUTH_ENABLED=true	JWT-based authentication with username/password

Authentication Endpoints (when enabled)

Endpoint	Method	Description
/auth/login	POST	Login with username/password, returns JWT tokens
/auth/refresh	POST	Refresh access token using refresh token
/auth/logout	POST	Logout (client discards tokens)
/auth/me	GET	Get current user information
/auth/switch-workspace	POST	Switch active workspace

Roles & Permissions

Role	Level	Permissions
Viewer	0	Read-only access to all data
Developer	1	+ Create/edit environments, schedules, run dbt commands
Admin	2	+ Manage users, plugins, workspaces, global settings

RBAC by Feature

Feature	Viewer	Developer	Admin
View models, lineage, catalog	✅	✅	✅
View runs and history	✅	✅	✅
Execute dbt commands	❌	✅	✅
Create/edit environments	❌	✅	✅
Create/edit schedules	❌	✅	✅
Enable/disable plugins	❌	❌	✅
Manage users	❌	❌	✅
Manage workspaces	❌	❌	✅

---

⚙️ Environment Variables

Database Settings

Variable	Default	Description
POSTGRES_HOST	localhost	PostgreSQL server hostname
POSTGRES_PORT	5432	PostgreSQL server port
POSTGRES_USER	user	PostgreSQL username
POSTGRES_PASSWORD	password	PostgreSQL password
POSTGRES_DB	dbt_workbench	PostgreSQL database name
DATABASE_URL	-	Override full database URL (optional)

Core Application

Variable	Default	Description
BACKEND_PORT	8000	Backend API server port
DBT_ARTIFACTS_PATH	./dbt_artifacts	Path to dbt artifacts directory
DBT_PROJECT_PATH	./dbt_project	Path to dbt project for execution
GIT_REPOS_BASE_PATH	./data/repos	Base path for cloned Git repositories

Authentication & Security

Variable	Default	Description
AUTH_ENABLED	false	Enable JWT authentication
SINGLE_PROJECT_MODE	true	Single workspace mode (no workspace switching)
JWT_SECRET_KEY	change_me	CHANGE IN PRODUCTION - Secret key for JWT signing
JWT_ALGORITHM	HS256	JWT signing algorithm
ACCESS_TOKEN_EXPIRE_MINUTES	30	Access token expiration time
REFRESH_TOKEN_EXPIRE_MINUTES	43200	Refresh token expiration (30 days)

Password Policy

Variable	Default	Description
PASSWORD_MIN_LENGTH	12	Minimum password length
PASSWORD_REQUIRE_UPPERCASE	true	Require uppercase letter
PASSWORD_REQUIRE_LOWERCASE	true	Require lowercase letter
PASSWORD_REQUIRE_NUMBER	true	Require number
PASSWORD_REQUIRE_SPECIAL	false	Require special character

Default Workspace

Variable	Default	Description
DEFAULT_WORKSPACE_KEY	default	Default workspace key identifier
DEFAULT_WORKSPACE_NAME	Default dbt Project	Default workspace display name
DEFAULT_WORKSPACE_DESCRIPTION	Default workspace	Default workspace description

Artifact Watcher

Variable	Default	Description
ARTIFACT_POLLING_INTERVAL	5	Polling interval in seconds
MAX_ARTIFACT_VERSIONS	10	Maximum artifact versions to retain
MONITORED_ARTIFACT_FILES	manifest.json,run_results.json,catalog.json	Files to monitor for changes

Lineage Configuration

Variable	Default	Description
DEFAULT_GROUPING_MODE	none	Default graph grouping (none, schema, tag)
MAX_INITIAL_LINEAGE_DEPTH	4	Maximum initial graph depth
LOAD_COLUMN_LINEAGE_BY_DEFAULT	false	Load column-level lineage by default
LINEAGE_PERFORMANCE_MODE	balanced	Performance mode (fast, balanced, detailed)

dbt Execution

Variable	Default	Description
MAX_CONCURRENT_RUNS	1	Maximum concurrent dbt runs
MAX_RUN_HISTORY	100	Maximum runs to keep in history
MAX_ARTIFACT_SETS	50	Maximum artifact sets to retain
LOG_BUFFER_SIZE	1000	Log buffer size in lines

Data Catalog

Variable	Default	Description
ALLOW_METADATA_EDITS	true	Allow editing catalog metadata
SEARCH_INDEXING_FREQUENCY_SECONDS	30	Search index refresh interval
FRESHNESS_THRESHOLD_OVERRIDE_MINUTES	-	Override source freshness threshold
VALIDATION_SEVERITY	warning	Default validation severity
STATISTICS_REFRESH_POLICY	on_artifact_change	When to refresh column statistics

Scheduler

Variable	Default	Description
SCHEDULER_ENABLED	true	Enable the scheduler background process
SCHEDULER_POLL_INTERVAL_SECONDS	30	Scheduler polling interval
SCHEDULER_MAX_CATCHUP_RUNS	10	Maximum catch-up runs on restart
SCHEDULER_DEFAULT_TIMEZONE	UTC	Default timezone for schedules

The scheduler normalizes all persisted timestamps to UTC before evaluating due runs. This prevents timezone-aware/naive comparison errors and keeps cron calculations consistent when databases return naive datetimes.

Scheduled runs inherit the dbt profile and target from their selected environment, ensuring the scheduled execution matches the active environment settings.

SQL Workspace

Variable	Default	Description
SQL_WORKSPACE_DEFAULT_CONNECTION_URL	-	Required - Database URL for SQL queries
SQL_WORKSPACE_MAX_ROWS	5000	Maximum rows returned per query
SQL_WORKSPACE_TIMEOUT_SECONDS	60	Query execution timeout
SQL_WORKSPACE_ALLOW_DESTRUCTIVE_DEFAULT	false	Allow destructive queries by default

The SQL Workspace supports two execution modes:

• Custom SQL: freeform queries against the selected environment, honoring destructive-query guardrails and row limits.
• dbt models: dual-pane editor showing the model source next to the compiled SQL (read-only). The compiled SQL is refreshed per environment/target and is the only code sent to the warehouse for execution. Viewer roles can inspect compiled SQL, while Developers/Admins can execute it.

Additional behaviors:

• Profiling is always enabled for executed queries so column statistics are consistently available alongside results.
• Results are paginated in the UI to keep browsing responsive; server-side limits (such as SQL_WORKSPACE_MAX_ROWS) still cap the dataset returned from the warehouse.

Notifications

Variable	Default	Description
NOTIFICATIONS_SLACK_TIMEOUT_SECONDS	10	Slack notification timeout
NOTIFICATIONS_WEBHOOK_TIMEOUT_SECONDS	10	Webhook notification timeout
NOTIFICATIONS_EMAIL_FROM	dbt-workbench@example.com	Email sender address
NOTIFICATIONS_EMAIL_SMTP_HOST	localhost	SMTP server host
NOTIFICATIONS_EMAIL_SMTP_PORT	25	SMTP server port
NOTIFICATIONS_EMAIL_USE_TLS	false	Use TLS for SMTP
NOTIFICATIONS_EMAIL_USERNAME	-	SMTP username
NOTIFICATIONS_EMAIL_PASSWORD	-	SMTP password

Plugin System

Variable	Default	Description
PLUGIN_SYSTEM_ENABLED	true	Enable the plugin system
PLUGINS_DIRECTORY	./plugins	Plugin discovery directory
PLUGIN_HOT_RELOAD_ENABLED	true	Enable hot-reload on file changes
PLUGIN_API_VERSION	1.0.0	Plugin API version
PLUGIN_ALLOWED_ENV_PREFIXES	DBT_,DBT_WORKBENCH_	Allowed environment variable prefixes for plugins

AI Copilot

Variable	Default	Description
AI_ENABLED	true	Enable AI Copilot endpoints and UI
AI_SECRETS_MASTER_KEY	-	Master key used to encrypt workspace AI secrets at rest
AI_DEFAULT_MODE	direct	Default provider mode (direct or mcp)
AI_DEFAULT_DIRECT_PROVIDER	openai	Default direct provider (openai, anthropic, gemini)
AI_DEFAULT_DIRECT_MODEL_OPENAI	-	Default OpenAI model override
AI_DEFAULT_DIRECT_MODEL_ANTHROPIC	-	Default Anthropic model override
AI_DEFAULT_DIRECT_MODEL_GEMINI	-	Default Gemini model override
AI_ALLOW_SESSION_PROVIDER_OVERRIDE	true	Allow per-session provider override in UI
AI_MCP_LOCAL_ALLOWLIST_JSON	{}	JSON allowlist for local MCP stdio templates
AI_MCP_PROCESS_IDLE_TTL_SECONDS	300	Idle timeout for local MCP processes
AI_MCP_CONNECT_TIMEOUT_SECONDS	20	Connection timeout for MCP calls
AI_MAX_INPUT_TOKENS	-	Optional soft max input tokens (no hard enforcement by default)
AI_MAX_OUTPUT_TOKENS	-	Optional soft max output tokens (no hard enforcement by default)
AI_AUDIT_RETENTION_DAYS	-	Optional AI audit retention window

---

🧩 Features Overview

Phase 1 — Artifact Viewer (Complete)

• Browse models, sources, tests
• Model details (columns, metadata)
• Basic lineage graph
• Runs list + statuses
• Dashboard overview

Phase 2 — Live Metadata Updates (Complete)

• Auto-detect changes to dbt artifacts
• Background watcher reloads metadata
• Frontend shows update indicators
• In-memory versioning

Phase 3 — dbt Execution Engine (Complete)

• Run dbt commands from UI
• Quick-launch Run/Test/Seed/Docs actions directly from the dbt Execution page via dedicated execution buttons (no separate command selector)
• Real-time log streaming
• Persist artifacts per run
• Dashboard surfaces the latest run status from execution history for consistency with the Run History page
• Seed-aware execution guardrails warn when seeds are required before running downstream models

Phase 4 — Metadata Persistence Layer (Complete)

• PostgreSQL backend
• Historical model snapshots

Phase 5 — Advanced Lineage (Complete)

• Column-level lineage derived from manifest and catalog artifacts
• Grouping by schema, resource type, and tags with collapsible aggregates
• Interactive expand/collapse of subgraphs to simplify large projects
• Upstream and downstream impact highlighting at model and column granularity
• Configurable defaults for grouping mode, graph depth, and column-level loading

Phase 6 — Scheduler (Complete)

• Cron-style scheduled runs with timezone support
• Email, Slack, and webhook notifications
• Environment-specific configurations
• Retry policies with exponential backoff
• Catch-up and overlap policies
• Historical run diagnostics with failure reasons, attempt timelines, and log/artifact links directly in the Schedules page
• Manual and cron-triggered runs expose queued/in-progress status with log links immediately after dispatch so "Run now" is easy to trace without waiting for completion
• Workspace-aware execution paths so scheduled dbt commands use the connected Git repository instead of the default project folder

Profile management in Environments (New)

• View configured dbt profiles and their available targets directly from the Environments page
• Add a new profile with a YAML snippet that is merged into profiles.yml
• Edit the full profiles.yml file inline when you need advanced changes

Phase 7 — SQL Workspace (Complete)

• SQL editor with syntax highlighting
• Dual-pane dbt model view showing editable source alongside read-only compiled SQL
• Query execution against configured database with dbt model runs using compiled SQL only, plus clear compile reminders when compiled artifacts are missing
• Environment-aware compilation and execution with role-based run restrictions
• Result profiling and statistics shared across custom SQL and dbt model runs
• Query history with execution mode, model references, and compiled SQL checksums

Projects & Version Control enhancements (New)

• Edit tracked files inline with save/commit controls and inline validation messaging
• Create new project files directly from the UI alongside the repository browser
• Browse project files in a hierarchical tree with folder expand/collapse and name filtering in Version Control and SQL Workspace

Phase 8 — Data Catalog Layer (Complete)

• Global fuzzy/prefix search across models, sources, exposures, macros, tests, tags, and columns
• Rich entity detail pages with dbt metadata, owners, tags, documentation, lineage previews, and column statistics
• Test health indicators surfaced in search, detail pages, and validation reports
• Source freshness visibility (max loaded timestamp, age, thresholds, status, last check)
• Persistent metadata enrichment for owners/tags/descriptions with optional edit controls
• Column-level descriptions, data types, nullability, and statistics synced from catalog.json
• Validation of missing documentation, owners/tags, failing tests, freshness gaps, and stale sources

Phase 9 — RBAC & Multi-Project (Complete)

• JWT-based authentication (optional)
• Role-based access control (Viewer, Developer, Admin)
• Multiple workspaces with independent data
• Workspace switching and per-user defaults

Phase 10 — Plugin Ecosystem (Complete)

• Backend plugin manager with manifest validation, capability/permission checks, and lifecycle events
• Hot-reloadable plugins discovered from the configurable PLUGINS_DIRECTORY (default ./plugins)
• Admin APIs to list, enable, disable, and reload plugins without restarting the server
• Workspace-scoped plugin configuration API for per-workspace settings
• Frontend marketplace and installed-plugins views with dynamic enable/disable controls
• Standardized plugin layout (/plugins/<name>/manifest.json, backend/, frontend/, static/)
• Safe opt-out via PLUGIN_SYSTEM_ENABLED=false for minimal installations

Phase 11 — Git-Integrated dbt Workspace (Complete)

• Workspace-scoped Git connections with branch switching, pull, push, and commit workflows
• In-app file tree with SQL/Jinja editor for models and YAML editor support for dbt configs
• Git-aware commit diffing, status, and history panels plus audit log visibility
• Role-aware editing controls for protected configuration files and conflict handling cues
• Model diff viewer
• Historical lineage browser

Phase 12 — AI Copilot (Complete)

• Global AI assistant panel with workspace-aware conversations
• Runtime provider switching between direct LLM APIs and MCP servers
• SQL copilot actions in SQL Workspace (explain/generate/optimize/fix prompts)
• Lineage Q&A and run troubleshooting context actions
• Admin-configurable AI defaults, secrets, and MCP server registry
• Proposal-based execution flow for SQL/dbt actions with explicit confirmation
• Full AI prompt/response/tool/action audit persistence per workspace

---

🔗 API Reference

Lineage API

Endpoint	Method	Description
/lineage/graph	GET	Model-level lineage with grouping metadata
/lineage/columns	GET	Column-level lineage graph
/lineage/model/{unique_id}	GET	Parents, children, and columns for a model
/lineage/upstream/{id}	GET	Upstream impact analysis
/lineage/downstream/{id}	GET	Downstream impact analysis
/lineage/groups	GET	Grouping metadata for schemas, types, tags

Catalog API

Endpoint	Method	Description
/catalog/entities	GET	List all catalog entities
/catalog/entities/{unique_id}	GET	Full entity detail with columns and tests
/catalog/search	GET	Fuzzy/prefix search across all entities
/catalog/validation	GET	Validation issues report
/catalog/entities/{unique_id}	PATCH	Update entity metadata (owner, tags, description)
/catalog/entities/{unique_id}/columns/{column_name}	PATCH	Update column-level metadata

Execution API

Endpoint	Method	Description
/execution/run	POST	Start a new dbt run
/execution/runs	GET	List run history
/execution/runs/{run_id}	GET	Get run details
/execution/runs/{run_id}/logs	GET	Stream run logs
/execution/runs/{run_id}/artifacts	GET	Get run artifacts
/execution/runs/{run_id}/cancel	POST	Cancel a running job

Run log streams now emit a terminal message even when the underlying dbt process fails before producing output. This ensures scheduler-triggered runs surface useful error context (for example, missing dbt installations) instead of showing an empty log window.

Scheduler API

Endpoint	Method	Description
/schedules	GET	List all schedules
/schedules	POST	Create a new schedule
/schedules/{id}	GET	Get schedule details
/schedules/{id}	PUT	Update schedule
/schedules/{id}	DELETE	Delete schedule
/schedules/{id}/pause	POST	Pause schedule
/schedules/{id}/resume	POST	Resume schedule
/schedules/{id}/run	POST	Trigger immediate run
/schedules/environments	GET	List environments
/schedules/environments	POST	Create environment

Plugin API

Endpoint	Method	Description
/plugins/installed	GET	List installed plugins
/plugins/{name}/enable	POST	Enable a plugin (Admin)
/plugins/{name}/disable	POST	Disable a plugin (Admin)
/plugins/reload	POST	Hot-reload plugins (Admin)
/plugins/config	GET	List workspace plugin configs
/plugins/config/{name}	GET	Get plugin config
/plugins/config	POST	Create plugin config (Admin)
/plugins/config/{name}	PUT	Update plugin config (Admin)
/plugins/config/{name}	DELETE	Delete plugin config (Admin)

AI API

Endpoint	Method	Description
/ai/settings	GET	Get effective AI workspace settings (sanitized)
/ai/settings	PUT	Update AI workspace defaults (Admin)
/ai/settings/secrets	PUT	Upsert encrypted AI secrets (Admin)
/ai/mcp/templates	GET	List local MCP allowlist templates (Admin)
/ai/mcp/servers	GET	List workspace MCP servers
/ai/mcp/servers	POST	Create MCP server config (Admin)
/ai/mcp/servers/{id}	PUT	Update MCP server config (Admin)
/ai/mcp/servers/{id}	DELETE	Delete MCP server config (Admin)
/ai/conversations	GET	List AI conversations in current workspace
/ai/conversations	POST	Create AI conversation
/ai/conversations/{id}/messages	GET	List conversation messages
/ai/chat/stream	POST	Stream AI response via SSE
/ai/actions/{proposal_id}	GET	Get proposal details
/ai/actions/{proposal_id}/confirm	POST	Confirm and execute proposal (Developer/Admin)
/ai/actions/{proposal_id}/reject	POST	Reject proposal

Git API

Endpoint	Method	Description
/git/connect	POST	Clone and connect a repository
/git/status	GET	Get repository status
/git/branches	GET	List branches
/git/checkout	POST	Switch branch
/git/pull	POST	Pull latest changes
/git/push	POST	Push commits
/git/commit	POST	Create a commit
/git/history	GET	Get commit history
/git/diff	GET	Get file diff
/git/files	GET	List repository files
/git/files/{path}	GET	Read file content
/git/files/{path}	PUT	Write file content

Configuration API

Endpoint	Method	Description
/config	GET	Get application configuration
/workspaces/active	GET	Get active workspace
/workspaces	GET	List workspaces (when auth enabled)

---

🧪 Testing

Backend

cd backend
pytest

Frontend

cd frontend
npm test

Documentation Site (Docusaurus)

cd website
npm test

Frontend CI note: GitHub Actions caches npm dependencies from frontend/package-lock.json. If you add or update UI dependencies, commit the updated lockfile in frontend/ so the actions/setup-node step can restore the cache and avoid lockfile-missing failures.

Docs CI note: The GitHub Pages workflow caches npm dependencies from website/package-lock.json. If you update docs site dependencies, commit the updated lockfile in website/ so the Pages build can restore the cache and run npm ci successfully.

Unit and integration coverage uses Vitest with React Testing Library (jsdom). The suite exercises the Version Control workflows (connected vs. missing repositories), dashboard fallbacks, and plugin pages. End-to-end UI checks use Playwright:

cd frontend
npm run test:e2e

---

📚 Documentation

• Architecture: ARCHITECTURE.md
• Plugin system: PLUGIN_SYSTEM.md
• Contributing: CONTRIBUTING.md
• Roadmap: ROADMAP.md
• Docs website: website/ (Docusaurus source + build scripts)

---

🌐 Why Choose dbt-Workbench?

For dbt Lineage & Visualization

• Interactive DAGs: Explore model dependencies with deterministic D3/dagre layouts
• Column-Level Lineage: Trace data flow from sources to marts at column granularity
• Impact Analysis: Instantly see upstream and downstream impacts of changes
• Open Source: No vendor lock-in, full customization capability

For dbt Run Orchestration

• Web-Based Execution: Run dbt commands directly from the browser
• Real-Time Logs: Stream logs as commands execute
• Run History: Track all executions with full context
• Cron Scheduling: Automate runs with timezone-aware scheduling

For Data Teams

• Self-Hosted: Keep all data within your infrastructure
• Air-Gapped Support: Run completely offline in secure environments
• Multi-Project: Manage multiple dbt projects in isolated workspaces
• RBAC: Role-based access control (Viewer, Developer, Admin)

For Organizations

• Cost Effective: No per-seat or per-run pricing
• Compliance Ready: Audit logging, encryption, access controls
• Extensible: Plugin system for custom integrations
• Enterprise Features: JWT auth, workspace isolation, Git integration

🔍 SEO Keywords

dbt-Workbench is designed for teams searching for:

• Open source dbt UI alternatives to dbt Cloud
• Self-hosted dbt lineage tools for data governance
• Air-gapped dbt deployment options for regulated industries
• dbt catalog and documentation viewers for data discovery
• SQL workspace for dbt projects for analytics engineering
• Data lineage visualization tools for impact analysis

🤝 Contributing

Contributions are welcome! See CONTRIBUTING.md for style guidelines, workflows, and expectations.

---

📜 License

MIT License — fully permissive for commercial and open-source use.

---

⭐ Star History

⭐ Support

If dbt-Workbench helps you, please star the repository to support the project.