How digital-ai/release-integration-template-python works
Project Type and Purpose
• This is a container-based plugin template for Digital.ai Release.
• It provides Python-based “tasks” (hello, sample_release_api_task, sample_server_task) packaged as Docker images.
• A local dev environment spins up:
• a Release server
• a Docker registry
• a remote-runner service
File Structure
• Top-level
• build.sh / build.bat – build & publish plugin artifacts (zip + Docker image)
• project.properties – plugin metadata (name, version)
• requirements.txt – Python dependencies
• .xebialabs/config.yaml – Release server connection & upload settings
• dev-environment/
• docker-compose.yaml – defines three services:
– digitalai-release (the server)
– digitalai-release-remote-runner (executes containers)
– digitalai-release-setup (initial configuration)
• Each service has its own Dockerfile
• src/
• Python task modules implementing the plugin’s logic
• resources/
• plugin-version.properties, type-definitions.yaml – type schemas & versioning
• tests/
• unit/integration tests for your tasks
Extract Key Architectural Insights from README
• The dev environment uses Docker Compose to stand up a full Release stack.
• build.sh/upload scripts drive the CI flow:
1. package code → zip
2. docker build → push to the registry
3. (optionally) upload zip to the Release server via SDK/CLI
Main Components
a. Developer Workstation (IDE, build scripts)
b. Local Docker Registry (container-registry)
c. Digital.ai Release Server (docker-compose service)
d. Remote-Runner Service (executes plugin containers)
e. Plugin Container (your Python code inside a Docker image)
f. Plugin ZIP Artifact (uploaded to server via .xebialabs config)
Relationships & Data Flows
• Developer triggers build.sh / build.bat →
• Packages code into ZIP →
– Pushes ZIP to Release Server (via SDK CLI, using .xebialabs/config.yaml)
• Builds Docker image →
– Tags & pushes image to the local registry (container-registry:5000)
• Release Server UI / API:
• Lists available plugins (from uploaded ZIP)
• When a release pipeline runs:
– Server posts a “run task” request to Remote-Runner
• Remote-Runner pulls the plugin image from registry →
• Starts the plugin container →
– Container executes your Python task (hello.py, sample_*.py)
– Task uses the Release SDK to call back into the server’s REST API
– Logs & results flow back to the Release server UI
Architectural Patterns & Principles
• Microservices / Containerization: each service in its own Docker container
• Separation of Concerns:
• Build & packaging layer (developer scripts)
• Orchestration layer (docker-compose, Release server)
• Execution layer (remote-runner + plugin container)
• Configuration as Code: docker-compose.yaml, .xebialabs/config.yaml, project.properties
Critical Elements
• Component Names matching folder names (dev-environment/digitalai-release, remote-runner, src module names)
• Artifacts: plugin ZIP, Docker image, test reports
• Data flows with arrowheads and labels