Skip to Content

Methods Reference

Public API methods for controlling workflows, generating content, and managing AI operations.

On This Page

Workflow Control

action_start_workflow()

Initiates workflow execution for a project.

Python 
def action_start_workflow(self):
    """
    Start the AI workflow for this project.

    Requires:
        - ai_template_id must be set
        - ai_state must be 'draft'

    Returns:
        bool: True if started successfully

    Raises:
        UserError: If prerequisites not met
    """

# Usage
project = env['project.project'].browse(1)
project.action_start_workflow()

action_next_stage()

Advances the project to the next workflow stage.

Python 
def action_next_stage(self):
    """
    Move project to the next stage in sequence.

    The next stage is determined by:
    1. Explicit next_stage_id on current stage
    2. Next stage by sequence number
    3. Final stage if none remaining

    Returns:
        ai.project.stage: The new current stage
    """

# Usage
next_stage = project.action_next_stage()

action_pause_workflow()

Pauses workflow execution.

Python 
def action_pause_workflow(self):
    """
    Pause the workflow execution.

    - Sets ai_state to 'paused'
    - Preserves current stage position
    - Can be resumed with action_resume_workflow()
    """

project.action_pause_workflow()

Content Generation

generate_content(template, context)

Generate content using an AI template.

Python 
def generate_content(self, template, context=None):
    """
    Generate content using AI provider and template.

    Args:
        template (ai.template): The prompt template to use
        context (dict): Variables for template substitution

    Returns:
        str: Generated content

    Example:
        template = env.ref('ai_workflow_website.page_template')
        content = project.generate_content(template, {
            'page_type': 'about',
            'company_name': 'Acme Inc',
        })
    """

process_task(task)

Process a single task through AI generation.

Python 
def process_task(self, task):
    """
    Process a task using the stage's AI template.

    Args:
        task (project.task): Task to process

    Effects:
        - Sets task.ai_state to 'processing'
        - Calls AI provider
        - Stores result in task.ai_content
        - Sets task.ai_state to 'done' or 'failed'

    Returns:
        bool: True if successful
    """

Approval Methods

action_approve_stage()

Approve the current stage and continue.

Python 
def action_approve_stage(self):
    """
    Approve the current stage content.

    Requirements:
        - Current stage must have requires_approval=True
        - ai_state must be 'approval_pending'

    Effects:
        - Records approval timestamp
        - Advances to next stage
        - Continues workflow execution
    """

project.action_approve_stage()

action_request_changes(feedback)

Request changes with feedback for regeneration.

Python 
def action_request_changes(self, feedback):
    """
    Request changes to the current stage output.

    Args:
        feedback (str): Instructions for what to change

    Effects:
        - Stores feedback
        - Re-runs the previous AI stage with feedback context
        - Returns to approval_pending when complete
    """

project.action_request_changes(
    feedback="Make the headline more compelling and add a testimonial section"
)

Utilities

get_workflow_progress()

Get workflow completion percentage.

Python 
@api.depends('stage_id', 'ai_template_id')
def _compute_ai_progress(self):
    """
    Compute workflow progress as percentage.

    Returns:
        int: Progress 0-100 based on stage position
    """

# Access via field
progress = project.ai_progress  # e.g., 45

get_stage_history()

Get the history of stage transitions.

Python 
def get_stage_history(self):
    """
    Get chronological list of stage transitions.

    Returns:
        list: [{
            'stage_id': int,
            'stage_name': str,
            'entered_at': datetime,
            'exited_at': datetime,
            'duration_seconds': int,
            'outcome': 'completed' | 'skipped' | 'failed'
        }]
    """

history = project.get_stage_history()