workflowforge 1.1.3

A robust and flexible library for creating GitHub Actions workflows, Jenkins pipelines, and AWS CodeBuild BuildSpecs programmatically in Python

WorkflowForge 🔨

📊 Project Status

A robust and flexible library for creating GitHub Actions workflows, Azure DevOps pipelines, Jenkins pipelines, and AWS CodeBuild BuildSpecs programmatically in Python.

✨ Features

• Intuitive API: Fluent and easy-to-use syntax
• Type Validation: Built on Pydantic for automatic validation
• IDE Support: Full autocompletion with type hints
• Type Safety: Complete mypy compliance with strict type checking
• Multi-Platform: GitHub Actions, Azure DevOps, Jenkins, AWS CodeBuild
• Pipeline Visualization: Automatic diagram generation with Graphviz
• Secrets Support: Secure credential handling across all platforms
• Templates: Pre-built workflows for common use cases
• Validation: Schema validation and best practices checking
• Optional Security Scan: On-demand Checkov scan for generated workflows (GitHub Actions and Azure)
• Optional AI Documentation: AI-powered README generation with OllamaPipelines)

🚀 Installation

pip install workflowforge

📚 Examples

Check out the examples/ directory for complete working examples:

# Run individual examples
python examples/github_actions/basic_ci.py
python examples/jenkins/maven_build.py
python examples/codebuild/node_app.py
python examples/azure_devops/hello_world.py
python examples/azure_devops/python_ci.py

This will generate actual pipeline files and diagrams using the new import structure.

🤖 AI Documentation (Optional)

WorkflowForge can automatically generate comprehensive README documentation for your workflows using Ollama (free local AI):

# Install Ollama (one-time setup)
curl -fsSL https://ollama.com/install.sh | sh
ollama serve
ollama pull llama3.2

# Generate workflow with AI documentation and diagram
workflow.save(".github/workflows/ci.yml", generate_readme=True, use_ai=True, generate_diagram=True)
# Creates: ci.yml + ci_README.md + CI_Pipeline_workflow.png

# Or generate README separately
readme = workflow.generate_readme(use_ai=True, ai_model="llama3.2")
print(readme)

Features:

• ✅ Completely free - no API keys or cloud services
• ✅ Works offline - local AI processing
• ✅ Optional - gracefully falls back to templates if Ollama not available
• ✅ Comprehensive - explains purpose, triggers, jobs, secrets, setup instructions
• ✅ All platforms - GitHub Actions, Azure DevOps, Jenkins, AWS CodeBuild

📊 Pipeline Visualization (Automatic)

WorkflowForge automatically generates visual diagrams of your pipelines using Graphviz:

# Install Graphviz (one-time setup)
brew install graphviz          # macOS
sudo apt-get install graphviz  # Ubuntu
choco install graphviz         # Windows

# Generate workflow with automatic diagram
workflow.save(".github/workflows/ci.yml", generate_diagram=True)
# Creates: ci.yml + CI_Pipeline_workflow.png

# Generate diagram separately
diagram_path = workflow.generate_diagram("png")
print(f"📊 Diagram saved: {diagram_path}")

# Multiple formats supported
workflow.generate_diagram("svg")  # Vector graphics
workflow.generate_diagram("pdf")  # PDF document

Features:

• ✅ Automatic generation - every pipeline gets a visual diagram
• ✅ Multiple formats - PNG, SVG, PDF, DOT
• ✅ Smart fallback - DOT files if Graphviz not installed
• ✅ Platform-specific styling - Azure DevOps (blue), GitHub (purple), Jenkins (orange), CodeBuild (toasted AWS yellow)
• ✅ Comprehensive view - shows triggers, jobs, dependencies, step counts

📖 Basic Usage

GitHub Actions Usage

from workflowforge import github_actions

# Create workflow using snake_case functions
workflow = github_actions.workflow(
    name="My Workflow",
    on=github_actions.on_push(branches=["main"])
)

# Create job
job = github_actions.job(runs_on="ubuntu-latest")
job.add_step(github_actions.action("actions/checkout@v4", name="Checkout"))
job.add_step(github_actions.run("echo 'Hello World!'", name="Say Hello"))

# Add job to workflow
workflow.add_job("hello", job)

# Generate YAML
print(workflow.to_yaml())

# Save with documentation and diagram
workflow.save(".github/workflows/hello.yml", generate_readme=True, generate_diagram=True)
# Creates: hello.yml + hello_README.md + My_Workflow.png

Jenkins Pipeline Usage

from workflowforge import jenkins_platform

# Create Jenkins pipeline using snake_case
pipeline = jenkins_platform.pipeline()
pipeline.set_agent(jenkins_platform.agent_docker("maven:3.9.3-eclipse-temurin-17"))

# Add stages
build_stage = jenkins_platform.stage("Build")
build_stage.add_step("mvn clean compile")
pipeline.add_stage(build_stage)

# Generate Jenkinsfile with diagram
pipeline.save("Jenkinsfile", generate_diagram=True)
# Creates: Jenkinsfile + jenkins_pipeline.png

AWS CodeBuild BuildSpec Usage

from workflowforge import aws_codebuild

# Create BuildSpec using snake_case
spec = aws_codebuild.buildspec()

# Add environment
env = aws_codebuild.environment()
env.add_variable("JAVA_HOME", "/usr/lib/jvm/java-17-openjdk")
spec.set_env(env)

# Add build phase
build_phase = aws_codebuild.phase()
build_phase.add_command("mvn clean package")
spec.set_build_phase(build_phase)

# Set artifacts
artifacts_obj = aws_codebuild.artifacts(["target/*.jar"])
spec.set_artifacts(artifacts_obj)

# Generate buildspec.yml with AI documentation and diagram
spec.save("buildspec.yml", generate_readme=True, use_ai=True, generate_diagram=True)
# Creates: buildspec.yml + buildspec_README.md + codebuild_spec.png

Azure DevOps Usage

Generate a simple “hello” pipeline:

from workflowforge import azure_devops as ado

pipeline = ado.hello_world_template_azure(
    name="Hello ADO",
    message="Hello Azure DevOps from WorkflowForge!",
)
pipeline.save("azure-pipelines.yml")

Or generate a Python matrix CI with caching across Ubuntu/Windows/macOS:

from workflowforge import azure_devops as ado

pipeline = ado.python_ci_template_azure(
    python_versions=["3.11", "3.12", "3.13"],
    os_list=["ubuntu-latest", "windows-latest", "macOS-latest"],
    use_cache=True,
)
pipeline.save("azure-pipelines.yml")

Optionally, scan the emitted YAML with Checkov (if installed):

from workflowforge import azure_devops as ado

pipeline = ado.python_ci_template_azure()
pipeline.save("azure-pipelines.yml", scan_with_checkov=True)

See also: examples/azure_devops/python_ci_scan.py.

AI Documentation Examples

# GitHub Actions with AI README
workflow = Workflow(name="CI Pipeline", on=on_push())
job = Job(runs_on="ubuntu-latest")
job.add_step(action("actions/checkout@v4"))
workflow.add_job("test", job)

# Save with AI documentation and diagram
workflow.save("ci.yml", generate_readme=True, use_ai=True, generate_diagram=True)
# Creates: ci.yml + ci_README.md + Test_Workflow.png

# Jenkins with AI README and diagram
pipeline = pipeline()
stage_build = stage("Build")
stage_build.add_step("mvn clean package")
pipeline.add_stage(stage_build)

# Save with AI documentation and diagram
pipeline.save("Jenkinsfile", generate_readme=True, use_ai=True, generate_diagram=True)
# Creates: Jenkinsfile + Jenkinsfile_README.md + jenkins_pipeline.png

# Check AI availability
from workflowforge import OllamaClient
client = OllamaClient()
if client.is_available():
    print("AI documentation available!")
else:
    print("Using template documentation (Ollama not running)")

Modular Import Structure

WorkflowForge supports platform-specific imports with snake_case naming following Python conventions:

Platform Modules

# Import specific platforms
from workflowforge import github_actions, jenkins_platform, aws_codebuild, azure_devops

# Or use short aliases
from workflowforge import github_actions as gh
from workflowforge import jenkins_platform as jenkins
from workflowforge import aws_codebuild as cb
from workflowforge import azure_devops as ado

Benefits

✅ Platform separation - Clear namespace for each platform ✅ Snake case naming - Follows Python PEP 8 conventions ✅ IDE autocompletion - Better IntelliSense support ✅ Shorter code - gh.action() vs github_actions.action()

🔧 Advanced Examples

Build Matrix Workflow

from workflowforge import github_actions as gh

job = gh.job(
    runs_on="ubuntu-latest",
    strategy=gh.strategy(
        matrix=gh.matrix(
            python_version=["3.11", "3.12", "3.13"],
            os=["ubuntu-latest", "windows-latest"]
        )
    )
)

Multiple Triggers

from workflowforge import github_actions as gh

workflow = gh.workflow(
    name="CI/CD",
    on=[
        gh.on_push(branches=["main"]),
        gh.on_pull_request(branches=["main"]),
        gh.on_schedule("0 2 * * *")  # Daily at 2 AM
    ]
)

Jobs with Dependencies

from workflowforge import github_actions as gh

test_job = gh.job(runs_on="ubuntu-latest")
deploy_job = gh.job(runs_on="ubuntu-latest")
deploy_job.needs = "test"

workflow = gh.workflow(name="CI/CD")
workflow.add_job("test", test_job)
workflow.add_job("deploy", deploy_job)

📚 Complete Documentation

Platform Support

GitHub Actions:

• on_push(), on_pull_request(), on_schedule(), on_workflow_dispatch()
• action(), run() steps
• secret(), variable(), github_context() for credentials
• Build matrices, strategies, environments
• Optional Checkov scan: workflow.save(path, scan_with_checkov=True)

Jenkins:

• pipeline(), stage(), agent_docker(), agent_any()
• jenkins_credential(), jenkins_env(), jenkins_param()
• Shared libraries, parameters, post actions

AWS CodeBuild:

• buildspec(), phase(), environment(), artifacts()
• codebuild_secret(), codebuild_parameter(), codebuild_env()
• Runtime versions, caching, reports

Azure DevOps:

• pipeline(), job(), strategy(matrix=...), task(), script()
• Build matrices, multi-OS matrix, and pip caching
• Hello world template
• Optional Checkov scan: pipeline.save(path, scan_with_checkov=True)

AI Documentation

• Ollama Integration: Local AI models (llama3.2, codellama, qwen2.5-coder)
• Automatic README: Explains workflow purpose, triggers, jobs, setup
• Fallback Support: Template-based documentation if AI unavailable
• All Platforms: Works with GitHub Actions, Azure DevOps, Jenkins, CodeBuild

Pipeline Visualization

• Graphviz Integration: Native diagram generation using DOT language
• Multiple Formats: PNG, SVG, PDF, DOT files
• Platform Styling: Color-coded diagrams (Azure DevOps: blue, GitHub: purple, Jenkins: orange, CodeBuild: toasted AWS yellow)
• Smart Fallback: DOT files if Graphviz not installed, images if available
• Comprehensive View: Shows triggers, jobs, dependencies, step counts, execution flow

🤝 Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Add tests
4. Submit a pull request

👨‍💻 Author & Maintainer

Brainy Nimbus, LLC - We love opensource! 💖

Website: brainynimbus.io Email: [e-mail] GitHub: @brainynimbus

📄 License

MIT License - see LICENSE for details.

🔗 Links

GitHub Actions:

• GitHub Actions Documentation
• Actions Marketplace
• Workflow Syntax

Jenkins:

• Jenkins Documentation
• Pipeline Syntax
• Jenkins Plugins

AWS CodeBuild:

• CodeBuild Documentation
• BuildSpec Reference
• CodeBuild Samples

Azure DevOps Pipelines:

• YAML schema reference
• Customize Python pipelines