dagucloud/dagu

Self-hostable workflow orchestrator for teams whose main work isn't orchestration. Declarative YAML over your scripts, SSH commands, containers, etc; keep workflows separate from business logic. One binary, no database, runs on limited H/W resources. Alternative to Airflow / Cron / Job Scheduler.

What it solves

Dagu is a local-first workflow engine designed for teams that need to orchestrate existing scripts, containers, and runbooks without the overhead of managing a complex platform. It replaces opaque cron jobs and heavy orchestrators like Airflow or Temporal by treating workflow structure as configuration rather than code, allowing users to add scheduling, retries, and observability to existing automation without modifying the underlying scripts.

How it works

Users define Directed Acyclic Graphs (DAGs) in a declarative YAML format. Dagu is distributed as a single binary that includes a Web UI, scheduler, and coordinator. It can run in three deployment models: as a single server, using temporary workers (e.g., via Kubernetes Jobs), or as distributed workers that poll a coordinator over gRPC. It natively supports executing shell commands, Docker containers, Kubernetes Jobs, and remote commands via SSH.

Who it’s for

It is intended for operations and internal automation teams who want to turn their existing scripts and runbooks into production workflows with minimal setup and operational burden.

Highlights

  • Single binary installation: No external database or message broker required.
  • Language-agnostic: Runs shell commands, Docker, K8s Jobs, and SSH without requiring scripts to import a framework.
  • Declarative YAML: Workflow structure, dependencies, and schedules are defined in YAML files.
  • Observability: Built-in Web UI for status tracking, logs, and visualization.
  • Durable Execution: Supports retries, overlap policies, and catch-up windows for scheduled runs.
  • Human Tasks: Ability to pause workflows for operator acknowledgement or input.
  • MCP Server: Built-in Model Context Protocol server for inspecting and controlling workflows via AI clients.

Related

  • Project
    argoproj/argo-workflows

    A container-native workflow engine for Kubernetes that orchestrates parallel jobs, specifically optimized for machine learning pipelines and data processing.

  • Project
    microsoft/conductor

    A CLI tool for defining and running deterministic, version-controlled multi-agent workflows using YAML configurations and the GitHub Copilot SDK.

  • Project
    apache/hamilton

    A lightweight Python library for creating portable and expressive data transformation DAGs, used to structure ML workflows, ETL pipelines, and RAG systems.

  • Project
    TangleML/tangle

    A visual orchestration system for ML experiments and production pipelines that allows users to build and run containerized batch workflows locally or in the cloud.

  • Project
    heymrun/heym

    An AI-native workflow automation platform for building, visualizing, and running multi-agent orchestrations and RAG pipelines via a no-code visual canvas.