What problem does it solve? Managing the lifecycle of harness learning proposals—inspecting tier distribution, fetching pending auto-update payloads, applying approved changes, and rolling back snapshots—requires a consistent bridge between the moai harness CLI and the orchestrator's user-approval flow, which this Skill provides. ## Core Features & Use Cases - Proposal Payload Production: Runs moai harness apply to fetch Tier 4 auto-update proposals as structured JSON payloads (proposal_id, target_path, field_key, new_value, confidence) for the orchestrator to surface via AskUserQuestion. - Apply/Rollback Orchestration: On approval, writes a decision file and executes the change through the 5-Layer Safety Pipeline; on rejection, removes the proposal; supports snapshot-based rollback via moai harness rollback <date>. - Learning Lifecycle Control: Checks tier distribution and rate-limit status with moai harness status, and disables learning with moai harness disable while preserving YAML comments. - Use Case: When the harness observer accumulates enough pattern observations to propose a skill description update, use this Skill to fetch the proposal, hand it to the orchestrator for user approval, and apply or discard it safely. ## Quick Start Ask the AI to check harness learning status and surface any pending auto-update proposals for approval.
