Dagu workflow orchestration: workflow authoring, web UI, and REST API
- version
- 0.1.10
- GitHub stars
- 27
Install
/plugin install dagu@vinnie357Run it inside Claude Code after adding the marketplace with /plugin marketplace add vinnie357/claude-skills.
Contents
Discovered from sourceSkills 3
Dagu REST API
Use this skill when integrating Dagu with external systems, automating workflow operations, or programmatically managing workflows through the API.
Core API Capabilities
The Dagu REST API provides endpoints for:
- Workflow Operations - Start, stop, retry workflows
- Status Queries - Get workflow and execution status
- DAG Management - List and inspect workflow definitions
- Execution History - Query past executions
- Log Retrieval - Fetch execution logs
Base URL
Default API base URL: http://localhost:8080/api/v1
Configure in Dagu settings if using a different host/port.
Quick Start Operations
Start a Workflow
POST /dags/{dagName}/start
Basic example:
curl -X POST http://localhost:8080/api/v1/dags/my_workflow/start
For parameter passing and advanced options, see references/api-endpoints.md.
Get Workflow Status
GET /dags/{dagName}/status
Returns current status, running steps, and execution details.
Stop a Workflow
POST /dags/{dagName}/stop
Stops currently running execution.
When to Consult References
Read references/api-endpoints.md for the complete endpoint reference: workflow operations (start/stop/retry/restart), status and history queries, log retrieval, DAG management, response schemas, HTTP status codes, and the error response format.
Common Use Cases
CI/CD Integration
Trigger Dagu workflows from your CI/CD pipeline:
# In GitHub Actions, GitLab CI, etc.
curl -X POST http://dagu-server:8080/api/v1/dags/deploy_production/start \
-H "Content-Type: application/json" \
-d '{"params": "VERSION=1.2.3 ENVIRONMENT=production"}'
Monitoring and Alerting
Query workflow status for external monitoring:
# Check if workflow is running
curl http://localhost:8080/api/v1/dags/critical_job/status
Build custom alerts based on status responses. See references/api-endpoints.md for response format details.
Dynamic Scheduling
Trigger workflows based on external events:
import requests
def trigger_workflow(dag_name, params=None):
url = f"http://localhost:8080/api/v1/dags/{dag_name}/start"
data = {"params": params} if params else {}
response = requests.post(url, json=data)
return response.json()
Response Formats
All API responses are JSON. Common response structure:
{
"status": "success",
"data": { ... }
}
Error responses:
{
"status": "error",
"message": "Error description"
}
For complete response schemas, consult references/api-endpoints.md.
Key Principles
- RESTful design: Standard HTTP methods (GET, POST, DELETE)
- JSON responses: All responses in JSON format
- Idempotent operations: Safe to retry most operations
- Error codes: Standard HTTP status codes
- Stateless: Each request is independent
Pro Tips
- Use the API for automation, use Web UI for manual operations
- Implement retry logic for network failures
- Cache DAG lists if querying frequently
- Use webhooks for event-driven workflows when possible
- Monitor API response times for performance issues
- Validate workflow names before calling API to avoid errors
Dagu Web UI
Use this skill when working with Dagu's web interface to manage workflows, view execution history, monitor running workflows, or configure the UI.
Core Capabilities
The Dagu Web UI provides:
- Workflow Management - View, start, stop, and manage workflows
- Execution Monitoring - Real-time status and logs
- History Viewing - Past execution records and results
- DAG Visualization - Visual representation of workflow structure
- Log Access - View detailed execution logs
- Schedule Management - Configure when workflows run
Quick Start
Access Dagu Web UI at http://localhost:8080 (default) after starting Dagu:
dagu server
Primary Operations
Start a Workflow
To manually execute a workflow:
- Navigate to workflow list
- Click the workflow name
- Click "Start" button
- View real-time execution progress
Monitor Execution
For detailed information on a running workflow, consult references/monitoring.md which covers:
- Reading execution logs
- Understanding status indicators
- Tracking step progress
- Identifying failures
View History
Execution history is not unconditionally preserved — the REST API supports deleting individual records (DELETE /dags/{dagName}/history/{requestId}, see /dagu:rest-api references/api-endpoints.md "Delete Execution History"). Open a workflow's history to review past runs, find failed executions, and retry them (see Common Tasks).
Workflow Visualization
The DAG view shows workflow structure. Click step names in the DAG view for step-specific details.
When to Consult References
Read references/monitoring.md for execution status indicators, real-time log viewing, progress tracking, error detection, and troubleshooting stuck workflows.
Common Tasks
Restart a Failed Workflow
- Find the failed execution in history
- Click the retry/restart button
- Monitor the new execution
Stop a Running Workflow
- Navigate to the running workflow
- Click "Stop" or "Cancel"
- Confirm the action
- View cleanup handlers execution
View Detailed Logs
When you need to debug a workflow:
- Click on the specific workflow execution
- Select the step with issues
- View stdout/stderr logs
- Check for error messages
For advanced log analysis, consult references/monitoring.md.
Key Principles
- Real-time visibility: Web UI provides live updates of workflow execution
- Click-based operations: No CLI needed for basic workflow management
- History preservation: Executions are logged and accessible; individual records can be deleted via the REST API
- Visual feedback: Status indicators show current state at a glance
- Log accessibility: Detailed logs available for debugging
Pro Tips
- Use the search feature to quickly find workflows by name
- Filter execution history by date range or status
- Click on step names in DAG view for step-specific details
- Use the refresh button if live updates seem delayed
- Check the scheduler status to verify cron jobs are active
Dagu Workflow Authoring
This skill activates when creating or modifying Dagu workflow definitions, configuring workflow steps, scheduling, or composing complex workflows.
Basic Workflow Structure
Minimal Workflow
# hello.yaml
steps:
- name: hello
command: echo "Hello from Dagu!"
Complete Workflow Structure
name: my_workflow
description: Description of what this workflow does
# Schedule (optional)
schedule: "0 2 * * *" # Cron format: daily at 2 AM
# Environment variables
env:
- KEY: value
- DB_HOST: localhost
# Parameters
params: ENVIRONMENT=production
# Email notifications (optional)
mailOn:
failure: true
success: false
smtp:
host: smtp.example.com
port: 587
errorMail:
from: dagu@example.com
to: alerts@example.com
# Workflow steps
steps:
- name: step1
command: echo "First step"
- name: step2
command: echo "Second step"
depends:
- step1
Steps
Basic Step
steps:
- name: greet
command: echo "Hello, World!"
Step with Script
steps:
- name: process
command: |
echo "Starting processing..."
./scripts/process.sh
echo "Done!"
Step with Working Directory
steps:
- name: build
dir: /path/to/project
command: make build
Step with Environment Variables
steps:
- name: deploy
env:
- ENVIRONMENT: production
- API_KEY: $API_KEY # From global env
command: ./deploy.sh
Executors
Command (default), Docker, SSH, HTTP, Mail, and JQ executor configuration: see references/executors.md.
Gotcha: the default executor (no executor: block) is the Command Executor — set
executor.type only when a step needs Docker, SSH, HTTP, Mail, or JQ.
Step Dependencies
Simple, multiple, and parallel dependency patterns via depends:: see references/steps-and-flow.md.
Gotcha: steps with no depends: run in parallel by default — sequencing is opt-in.
Conditional Execution
Preconditions
steps:
- name: deploy_production
preconditions:
- condition: "`echo $ENVIRONMENT`"
expected: "production"
command: ./deploy.sh
Continue On Failure
steps:
- name: optional_step
continueOn:
failure: true
command: ./might_fail.sh
- name: cleanup
depends:
- optional_step
command: ./cleanup.sh # Runs even if optional_step fails
Error Handling and Retries
Retry Configuration
steps:
- name: flaky_api_call
command: curl https://api.example.com/data
retryPolicy:
limit: 3
intervalSec: 10
Exponential Backoff
steps:
- name: with_backoff
command: ./external_api.sh
retryPolicy:
limit: 5
intervalSec: 5
exponentialBackoff: true # 5s, 10s, 20s, 40s, 80s
Signal on Stop
steps:
- name: graceful_shutdown
command: ./long_running_process.sh
signalOnStop: SIGTERM # Send SIGTERM instead of SIGKILL
Data Flow
Passing step output to later steps via output: and captured script: variables: see references/steps-and-flow.md.
Scheduling
Cron Schedule
# Daily at 2 AM
schedule: "0 2 * * *"
# Every Monday at 9 AM
schedule: "0 9 * * 1"
# Every 15 minutes
schedule: "*/15 * * * *"
# First day of month at midnight
schedule: "0 0 1 * *"
Start/Stop Times
# Only run during business hours
schedule:
start: "2024-01-01"
end: "2024-12-31"
cron: "0 9-17 * * 1-5" # Mon-Fri, 9 AM to 5 PM
Environment Variables
Global, step-level, and file-loaded (.env) environment variables: see references/steps-and-flow.md.
Parameters
Defining workflow params: and overriding them from the CLI: see references/steps-and-flow.md.
Sub-Workflows
Calling and nesting sub-workflows with run:: see references/steps-and-flow.md.
Handlers
Exit, failure, and success handlers via handlerOn:: see references/steps-and-flow.md.
Gotcha: handlerOn is declared once per workflow, alongside steps: — it is not nested inside an individual step.
Templates and Variables
Built-in ({{.Name}}, {{.Step.Name}}, {{.timestamp}}, {{.requestId}}) and custom params:-backed templates: see references/steps-and-flow.md.
Gotcha: custom parameters are referenced as {{.Params.NAME}}, not bare {{.NAME}} — the bare form is reserved for built-ins.
Common Patterns
Complete ETL pipeline, multi-environment deployment, database backup, and health-check monitoring workflow artifacts: see references/common-patterns.md.
Best Practices
Workflow Organization
# Good: Clear, descriptive names
name: user_data_sync
description: Synchronize user data from CRM to database
# Good: Logical step names
steps:
- name: fetch_from_crm
- name: validate_data
- name: update_database
# Avoid: Generic names
name: workflow1
steps:
- name: step1
- name: step2
Error Handling
# Always define error handlers for critical workflows
handlerOn:
failure:
- name: cleanup
command: ./cleanup.sh
- name: notify
executor:
type: mail
config:
to: team@example.com
# Use retries for flaky operations
steps:
- name: api_call
command: curl https://api.example.com
retryPolicy:
limit: 3
intervalSec: 5
exponentialBackoff: true
Environment Management
# Use parameters for environment-specific values
params: ENVIRONMENT=development
# Load environment from files
env:
- config/$ENVIRONMENT.env
# Override in production
# dagu start workflow.yaml ENVIRONMENT=production
Modular Workflows
# Break complex workflows into sub-workflows
steps:
- name: data_ingestion
run: workflows/ingestion.yaml
- name: data_transformation
run: workflows/transformation.yaml
depends:
- data_ingestion
Key Principles
- Keep workflows focused: One workflow per logical task
- Use dependencies wisely: Parallelize when possible
- Handle errors explicitly: Define failure handlers
- Use retries for flaky operations: Network calls, external APIs
- Parameterize configurations: Make workflows reusable
- Document workflows: Add clear names and descriptions
- Test workflows: Start with small, focused workflows
- Monitor and alert: Use handlers to track workflow health
References
references/executors.md— Command (default), Docker, SSH, HTTP, Mail, and JQ executor configurationreferences/steps-and-flow.md— step dependencies, data flow, environment variables, parameters, sub-workflows, handlers, templatingreferences/common-patterns.md— complete ETL, multi-environment deployment, backup, and monitoring workflow artifacts
Sponsors
Details
- Marketplace
- vinnie357/claude-skills
- Source
- plugins/tools/dagu
- Category
- Tools
- Keywords
- dagu, workflow, orchestration, automation
More from vinnie357/claude-skills
All 33- designv0.1.5
- elixirv0.1.25
- extras-claude-codev0.3.1
- githubv0.3.17
- graphifyv0.1.4
- linearv0.1.10
- playwrightv0.1.3
- pmv0.1.6
- proxmoxv0.1.1
- qav0.1.7
- rigv0.1.2
- runexv0.1.11
- rustv0.1.13
- sbxv0.1.3
- slidevv0.2.11
- tweagv0.1.5
- typescriptv0.1.1
- uiv0.1.7
- valev0.1.0
- wasmv0.1.11
- whammv0.1.4
- zigv0.1.11
- agent-sandboxingv0.2.15
- all-skillsv0.4.187
