OSOP Python SDK

Serves both SOP Doc and The Loop. Build workflow tools in Python.

Parse .osop files, validate, render, execute. Build custom SOP Doc viewers or integrate The Loop into Python applications. Sync and async clients.

Website: osop.ai | GitHub: github.com/osop/osop-sdk-py

Installation

pip install osop-sdk

Quick Start

from osop_sdk import OsopClient

client = OsopClient(base_url="http://localhost:8080")

# Validate a workflow
result = client.validate(file_path="deploy.osop.yaml")
print(result.valid)  # True

# Run a workflow in dry-run mode
execution = client.run(
    file_path="deploy.osop.yaml",
    inputs={"environment": "staging"},
    dry_run=True,
)
print(execution.status)  # "completed"

# Render a Mermaid diagram
diagram = client.render(file_path="deploy.osop.yaml", format="mermaid")
print(diagram.content)

# Run test cases
test_result = client.test(file_path="deploy.osop.yaml")
print(f"{test_result.passed}/{test_result.total} passed")

Async Usage

from osop_sdk import AsyncOsopClient

async def main():
    client = AsyncOsopClient(base_url="http://localhost:8080")
    result = await client.validate(file_path="workflow.osop.yaml")
    print(result.valid)

API Reference

OsopClient / AsyncOsopClient

Parameter	Type	Required	Description
base_url	str	Yes	OSOP server base URL
api_key	str	No	API key for authentication
timeout	float	No	Request timeout in seconds (default: 30.0)

Methods

Method	Description
validate(content=, file_path=, strict=)	Validate a workflow against the schema
run(content=, file_path=, inputs=, dry_run=, timeout_seconds=)	Execute a workflow
render(content=, file_path=, format=, direction=)	Render a workflow diagram
test(content=, file_path=, filter=, verbose=)	Run workflow test cases

Development

git clone https://github.com/osop/osop-sdk-py.git
cd osop-sdk-py
pip install -e ".[dev]"
pytest

License

Apache License 2.0 — see LICENSE for details.