AI agent
Asks for an approved operation without seeing private credentials.
Self-hosted · Single binary · Zero dependencies
The workflow engine that doesn't turn into an SRE project.
Dagu is a lightweight alternative to Airflow or Cron with a Web UI. Define DAGs in a simple declarative YAML format. It supports shell commands, docker containers, k8s jobs, remote commands via SSH, and more. It was designed to be easy to use, self-contained, and require no coding, making it ideal for small teams.
Dagu is built for developers who want powerful workflow orchestration without the operational overhead.
Linux
macOS
Windows
Docker
Kubernetes
dagu start-all --port 8080
Install dagu command
Self-contained
One binary, no external database or broker. Workflows run from files, so it works in air-gapped environments too.
CLI-oriented
Shell scripts, Python jobs, SQL, dbt, DuckDB, containers, and data runbooks all run as pipeline steps unchanged. You don't need a new framework.
Declarative
Define workflows in simple, declarative YAML format. You can use any existing script or tools without any modification.
Secure AI-agent integration
Use your favorite AI agent to create, improve, debug, and run workflows. Built-in secret management keeps credentials out of logs and away from agents, while still letting them preview and run tasks. Every action is logged for compliance.
Set up via MCP →workflow.yaml
# Controlled production runbook with typed inputs and review
name: "repair-customer-account"
params:
- name: customer_id
type: string
- name: change_scope
type: string
enum: ["metadata_only", "permissions"]
default: metadata_only
- name: dry_run
type: boolean
default: true
steps:
- id: inspect_account
run: ./scripts/inspect-account.sh --customer $${customer_id}
stdout:
artifact: reports/inspection.md
- id: review
action: noop
depends: [inspect_account]
approval:
prompt: "Review the inspection report before repair."
- id: repair_account
run: ./scripts/repair-account.sh --customer $${customer_id} --scope $${change_scope} --dry-run $${dry_run}
depends: [review]
stdout:
artifact: reports/repair.log
Trusted by developers at
NOW VIEWING · THE FLIGHT DECK
AI agent gateway
MCP gateway for running DAGs without exposing secrets
A built-in MCP server lets any AI agent manage your DAGs out of the box. Private systems sit behind a security boundary that requires secrets, and Dagu masks them automatically — so the agent never sees them.
Request
Request
Dagu MCP
Enforces granular access control before any workflow can run.
Authorize
Authorize
DAGs
Runs the exact script, container, SQL, HTTP, SSH, or sub-workflow you already trust.
Execute
Execute
Private systems
Production tools and data stay behind Dagu, not inside the agent context.
ShellSQLSSH
Agents request a named workflow action through Dagu MCP instead of receiving shell, database, or API credentials.
Dagu checks access, injects secrets only at runtime, and can pause sensitive workflows for human approval.
Every run keeps logs, outputs, status, history, and audit evidence for humans and agents to inspect.
Scoped access
Expose only the actions each agent is allowed to run.
Managed secrets
Inject credentials at runtime without revealing them to the agent.
Audit trail
Track who or what ran each workflow and with which inputs.
Approvals
Pause high-risk operations for human review.
Safe results
Return status and outputs without opening raw system access.
The agent gets a bounded tool. Your private systems keep the secrets, logs, permissions, and operational history in Dagu.
The control plane around your real work
Dagu gives teams the missing operational layer for scripts, tools, runbooks, and AI agents.
Observability
Track workflows, schedules, live status, logs, run history, outputs, artifacts, and approvals in the Web UI.
Language-agnostic execution
Run shell commands, Python scripts, containers, Kubernetes Jobs, SQL queries, HTTP requests, SSH commands, and more.
Reproducible runs
Pin tools in workflow YAML so workers install and cache the expected binaries before execution.
Multiplayer agents
Run shared AI agents inside workflows and expose workflows to agents as bounded, observable tools.
Built-in approvals
Pause sensitive steps for human review, confirmation, or intervention before execution continues.
MCP server
Let MCP-capable agents inspect Dagu state, preview changes, edit workflows, and control runs.
Harness-agnostic agents
Run coding agents and agent CLIs from workflows while keeping logs, approvals, secrets, and audit history in Dagu.
Secret management
Inject secrets into workflow runs and mask sensitive values in logs for users, agents, and chat tools.
Self-host or managed
Run a single binary, deploy with containers or Kubernetes, or use managed and hybrid deployment models.
Permission control
Use RBAC, SSO, API keys, workspace controls, and audit logs to govern workflow access.
Deployment models
Run a self-hosted instance, use the Dagu managed server, or combine the cloud control plane with private workers in your infrastructure.
Start on one machine. Scale to self-hosted, managed, or hybrid.
Local
Single server
Run `dagu start-all` on one machine with local file-backed state. No database, broker, or platform stack required.
Self-hosted
Your infrastructure
Keep the Dagu server, workers, secrets, logs, and execution inside your own environment.
Dagu
Full managed server
Use a dedicated Dagu server operated by Dagu in an isolated gVisor instance on GKE.
Hybrid
Managed server, private execution
Let Dagu operate the server while private workers run Docker, private-network, or data-local steps.
Hybrid execution
Managed Dagu server
Private worker over mTLS
Docker, private APIs, secrets, and data-local work
Hybrid keeps the Dagu server managed while execution that needs your network, runtime, or data stays under your control.
Production-grade workflow engine features
Battle-tested workflow engine features let you focus on the core value without worrying about orchestration.
Performance
Run thousands of workflow runs per day on one machine, depending on hardware, workflow shape, step duration, and queue settings.
Distributed executions
Use queues, concurrency limits, and distributed workers to control load and spread jobs across machines.
Notifications
Recurring schedules, catchup, durable executions, timeouts, reruns, notification, and incidents routing to responders.
Auth & RBAC
User management, RBAC, workspaces, approval steps, built-in secret management, API Keys, and webhooks.
Real-world use cases
01
Customer Support Automation
Turn repetitive engineering requests into self-service workflows. Business teams run them independently. Engineers stay out of the loop.
02
Cron and Legacy Script Management
Bring existing shell scripts, Python scripts, HTTP calls, and scheduled jobs into Dagu without rewriting them.
03
Infrastructure and Server Automation
Coordinate SSH backups, cleanup jobs, deploy scripts, patch windows, precondition checks, and lifecycle hooks.
04
ETL and Data Operations
Run PostgreSQL or SQLite queries, S3 transfers, jq transforms, validation steps, and reusable sub-workflows.
05
GitHub as the interface. Your infrastructure as the runtime.
Trigger Dagu runs from GitHub events from issues and PRs via managed Dagu GitHub App.
06
Container and Kubernetes Workflows
Compose workflows where each step can run a Docker image, Kubernetes Job, shell command, or validation step.
07
Media Conversion
Run ffmpeg, thumbnail extraction, audio normalization, image processing, and other compute-heavy jobs across workers.
08
IoT and Edge Workflows
Run sensor polling, local cleanup, offline sync, health checks, and device maintenance jobs on small devices.
09
AI Agent Workflows
Run any harness to automate certain development workflow with clean context and automatic retry & failover.
Common patterns
Health Check
SSH Backup
Notify
Private Script Workflows
Turn private scripts, internal CLIs, reusable actions, pinned tools, and artifacts into workflows other teams can safely request or run.
- 1Keep existing scripts and commands intact
- 2Expose narrow parameters instead of broad production access
- 3Call versioned Dagu Actions with typed outputs
- 4Store reports and logs as first-class artifacts
workflow.yaml
tools:
- jqlang/jq@jq-1.7.1
steps:
- id: collect
action: acme/dagu-action-export@v1.4.2
with:
dataset: customers
- id: transform
run: jq '.items[] | {id, email}' ${collect.outputs.path}
stdout:
artifact: reports/customers.json
depends: [collect]
- id: publish
action: outputs.write
with:
values:
report: reports/customers.json
rows: ${collect.outputs.rows}
depends: [transform]
Typed parameters
Dagu automatically renders them as a typed parameter input form in the Web UI, useful for making self-service workflows or operation by non-engineers.
workflow.yaml
tools:
- jqlang/jq@jq-1.7.1
steps:
- id: inspect
run: jq --version
- id: summarize
action: python-script@v1
with:
input:
rows: [42, 8]
script: |
return {"total": sum(input["rows"])}
Tool provisioning uses aqua as the default provider.
Pinned tools
Declare CLI tools and versions in YAML. Dagu prepares them before host-command steps run.
Learn moreOfficial and third-party actions
Use official Dagu Actions such as Python or NodeJS, or build your own action package to run your workflow with necessary tools and dependencies.
Learn moreArtifacts
Manage artifacts in DAG runs and preview / download in the Web UI.
Learn moreNotifications and incidents
Configure notifications for each workflow so necessary events reach the right channel.
Learn moreMultiplayer AI agent in Slack
AI agent in Slack helps you to manage your workflows,
debug failures, and recover incidents.
DaguMessage...
Full-fledged Workflow features
Dagu is equipped with production-grade workflow engine features.
Loved by developers
Real comments from the Dagu GitHub community.
This is literally my dream project. I have set up GitLab at some point just to run some commands and this project is perfect for all my use case. Thank you very much!
@vnghia
Developer
Dagu is really a gem in the open source world! ❤️ What a release — with GitHub Actions as well. Thanks guys.
@borestad
Elk Studios
Dagu is an amazing tool!! It works perfectly now, thank you so much! 💖
@jarnik
Freelance developer
I've been exploring Dagu for a few weeks and really love the Zero-Ops philosophy. The declarative YAML approach is exactly what the workflow orchestration space needs.
@vincent067
Developer
Amazing! I did not expect this to be delivered that fast! Multiple schedulers do not compete and the locking mechanism works as expected. Good job!
@jonasban
Developer
I've recently played around with Dagu for private stuff and think it could also be worth introducing at work.
@DarkWiiPlayer
Developer
Awesome, it works like a charm! I second the update of the doc on this one.
@thibmart1
Developer
That's very impressive. The exit code dictionary will make these DAGs even further sophisticated. Once again, very good features. 🫡
@ghansham
Developer
This is literally my dream project. I have set up GitLab at some point just to run some commands and this project is perfect for all my use case. Thank you very much!
@vnghia
Developer
Dagu is really a gem in the open source world! ❤️ What a release — with GitHub Actions as well. Thanks guys.
@borestad
Elk Studios
Dagu is an amazing tool!! It works perfectly now, thank you so much! 💖
@jarnik
Freelance developer
I've been exploring Dagu for a few weeks and really love the Zero-Ops philosophy. The declarative YAML approach is exactly what the workflow orchestration space needs.
@vincent067
Developer
Amazing! I did not expect this to be delivered that fast! Multiple schedulers do not compete and the locking mechanism works as expected. Good job!
@jonasban
Developer
I've recently played around with Dagu for private stuff and think it could also be worth introducing at work.
@DarkWiiPlayer
Developer
Awesome, it works like a charm! I second the update of the doc on this one.
@thibmart1
Developer
That's very impressive. The exit code dictionary will make these DAGs even further sophisticated. Once again, very good features. 🫡
@ghansham
Developer
This is literally my dream project. I have set up GitLab at some point just to run some commands and this project is perfect for all my use case. Thank you very much!
@vnghia
Developer
Dagu is really a gem in the open source world! ❤️ What a release — with GitHub Actions as well. Thanks guys.
@borestad
Elk Studios
Dagu is an amazing tool!! It works perfectly now, thank you so much! 💖
@jarnik
Freelance developer
I've been exploring Dagu for a few weeks and really love the Zero-Ops philosophy. The declarative YAML approach is exactly what the workflow orchestration space needs.
@vincent067
Developer
Amazing! I did not expect this to be delivered that fast! Multiple schedulers do not compete and the locking mechanism works as expected. Good job!
@jonasban
Developer
I've recently played around with Dagu for private stuff and think it could also be worth introducing at work.
@DarkWiiPlayer
Developer
Awesome, it works like a charm! I second the update of the doc on this one.
@thibmart1
Developer
That's very impressive. The exit code dictionary will make these DAGs even further sophisticated. Once again, very good features. 🫡
@ghansham
Developer
I love how simple this is now.
@n3storm
Estudio Nexos
Fantastic! Thanks for the quick reply — keep up the great project!
@mitchplze
Developer
Amazing support! Thank you!
@kacamific
Developer
It works perfectly with v2.6.1. Thank you very much for the correction.
@napnap75
Developer
Thanks for the fix. Works great now.
@triole
Developer
As always, appreciate the great work — just sharing feedback to help shape the roadmap.
@jonnochoo
Developer
it's awesome you keep working on it 👍
@fishnux
Developer
Thank you so much for the quick fix! I tested it, and it now correctly stores the files where it should.
@IngwiePhoenix
Developer
I love how simple this is now.
@n3storm
Estudio Nexos
Fantastic! Thanks for the quick reply — keep up the great project!
@mitchplze
Developer
Amazing support! Thank you!
@kacamific
Developer
It works perfectly with v2.6.1. Thank you very much for the correction.
@napnap75
Developer
Thanks for the fix. Works great now.
@triole
Developer
As always, appreciate the great work — just sharing feedback to help shape the roadmap.
@jonnochoo
Developer
it's awesome you keep working on it 👍
@fishnux
Developer
Thank you so much for the quick fix! I tested it, and it now correctly stores the files where it should.
@IngwiePhoenix
Developer
I love how simple this is now.
@n3storm
Estudio Nexos
Fantastic! Thanks for the quick reply — keep up the great project!
@mitchplze
Developer
Amazing support! Thank you!
@kacamific
Developer
It works perfectly with v2.6.1. Thank you very much for the correction.
@napnap75
Developer
Thanks for the fix. Works great now.
@triole
Developer
As always, appreciate the great work — just sharing feedback to help shape the roadmap.
@jonnochoo
Developer
it's awesome you keep working on it 👍
@fishnux
Developer
Thank you so much for the quick fix! I tested it, and it now correctly stores the files where it should.
@IngwiePhoenix
Developer
Quickstart
Install Dagu with the guided wizard, then continue in the full installation guide or quickstart docs.
1
Install dagu command
The script installers are the recommended path. Homebrew, npm, and Docker remain available for binary-only or container installs.
Mac/Linux Terminal
$curl -fsSL https://raw.githubusercontent.com/dagucloud/dagu/main/scripts/installer.sh | bash
✓ Guided installer ready
Optional
Connect AI tools
Install the Dagu skill for workflow authoring, or connect an MCP client to a running Dagu server.
Dagu skill
$gh skill install dagucloud/dagu dagu
Helps Claude Code, Codex, Gemini CLI, and other AI coding tools write Dagu YAML.
Dagu MCP endpoint
URLhttp://localhost:8080/mcp
Lets MCP-capable clients read Dagu state, preview changes, and control DAG runs.
OIDC/SSO-backed MCP access is available for enterprise deployments. Contact us to discuss setup.
2
Next steps
The guided installer can finish the first-run setup for you.
# What the installer can do
Add Dagu to your PATH
Set up a background service
Create and verify the first admin
Project Community
Discuss usage, report issues, and follow development.
Solution partner
Solution partner support for customer projects
For teams embedding Dagu into customer-facing products, internal platforms, or managed services, we provide enterprise features and hands-on integration support.
We help partners design the security model, integration path, and operating plan so Dagu can fit cleanly into customer projects without handing out broad production access.