Skip to content

Task Status Lifecycle

AgentWeave uses a structured task lifecycle to track work from creation to completion.

Status Flow

pending → assigned → in_progress → completed → under_review → approved
                                             ↘ revision_needed (loops back)
                                             ↘ rejected

Status Definitions

Status Meaning
pending Task created but not yet assigned
assigned Assigned to an agent, awaiting start
in_progress Agent is actively working on it
completed Work finished, awaiting review
under_review Being reviewed by assigner or principal
approved Review passed, task is done
revision_needed Changes required, loops back to in_progress
rejected Task rejected, will not be completed

Transitions

Typical happy-path transition:

agentweave task update <task_id> --status assigned
agentweave task update <task_id> --status in_progress
agentweave task update <task_id> --status completed
agentweave task update <task_id> --status approved

Requesting revision:

agentweave task update <task_id> --status revision_needed --note "Fix error handling in auth.py"

Task Structure

Tasks include:

  • title — short summary
  • description — detailed explanation
  • assignee — responsible agent
  • assigner — who created/assigned the task
  • prioritylow, medium, high, or critical
  • requirements — list of requirement strings
  • acceptance_criteria — list of criteria strings
  • deliverables — expected outputs