human-escalation

name	human-escalation
description	Define when and how the commander requests human input during the mothership workflow. Covers escalation triggers and escalation handling.

Human Escalation

Purpose

Define when and how commander requests human input.

Trigger Conditions

• requirement ambiguity
• QA cannot verify correctness
• risky or destructive change
• security or infrastructure impact
• unresolved product or design judgment
• cleanup anomaly requiring explicit approval

Expected Commander Output

A concise escalation should include:

• what happened
• why automation cannot proceed safely
• what decision is needed from the human
• current issue / PR / worktree state
• recommended options if available

Rule

Escalation is not failure. It is the correct action when confidence is insufficient.