Skip to main content

Working with Workflow

Workflow in Credo AI consists of two parts: Workflow Stages that define the governance pipeline, and Use Case Workflow that tracks each use case's progress through that pipeline.

Workflow Stages​

Workflow stages define the governance pipeline that use cases progress through. The system includes fixed start and end stages, with customizable intermediate stages.

Listing Workflow Stages​

from credoai import CredoAI

client = CredoAI()

response = client.workflow_stages.list()
for stage in response.items:
    print(f"{stage.name} (ID: {stage.id})")

Creating an Intermediate Stage​

Only intermediate workflow stages can be created via the API. Start and end stages are system-managed.

from credoai import WorkflowStageCreate

stage = client.workflow_stages.create(
    data=WorkflowStageCreate(
        name="Security Review",
    )
)
print(f"Created: {stage.id}")

Updating a Stage​

Start and end stages cannot be modified via the API.

from credoai import WorkflowStageUpdate

updated = client.workflow_stages.update(
    workflow_stage_id="ws_abc123",
    data=WorkflowStageUpdate(name="Updated Stage Name"),
)

Deleting a Stage​

Start and end stages cannot be deleted via the API.

client.workflow_stages.delete(workflow_stage_id="ws_abc123")

Use Case Workflow​

Each use case has a workflow that tracks its progress through the governance pipeline. The workflow has both a current stage and a current step within that stage.

Step progression within each stage: assessment -> evidence_collection -> clear

Getting Workflow Status​

use_case = client.use_cases("uc_abc123")

workflow = use_case.workflow.get()
print(f"Current stage: {workflow.stage}")
print(f"Current step: {workflow.step}")

Updating Workflow​

from credoai import UseCaseWorkflowUpdate

use_case = client.use_cases("uc_abc123")

updated = use_case.workflow.update(
    data=UseCaseWorkflowUpdate(stage_id="ws_abc123"),
)

Advancing the Step​

Move to the next step within the current workflow stage:

use_case = client.use_cases("uc_abc123")

# Advance: assessment -> evidence_collection -> clear
result = use_case.workflow.advance_step()
print(f"New step: {result.step}")

Advancing the Stage​

Move to the next workflow stage, resetting to the assessment step:

use_case = client.use_cases("uc_abc123")

result = use_case.workflow.advance_stage()
print(f"New stage: {result.stage}")

Rejecting a Use Case​

Set the use case's workflow step to rejected. This is a terminal state.

use_case = client.use_cases("uc_abc123")

result = use_case.workflow.reject()
print(f"Step: {result.step}")  # "rejected"

Next Steps​