autonomous-progress-and-escalation

What problem does it solve? Long-running tasks often stall because the assistant asks unnecessary confirmation questions or, conversely, proceeds unsafely past decisions that genuinely require user input. This Skill defines clear rules for when to continue autonomously and when to pause. ## Core Features & Use Cases - Communication Modes: Distinguishes progress updates, blocking questions, and final results so execution continues without waiting after non-blocking messages. - Escalation Criteria: Pauses only for genuine decisions, logins, sensitive confirmations, missing required material, destructive operations, or policy boundaries. - Recovery Pattern: Diagnoses failures, tries alternative safe methods, and escalates with a structured report only after recovery attempts are exhausted. - Use Case: During a multi-phase development task, the assistant fixes test failures and refactors code autonomously, sending brief milestone updates, and only stops to ask before deploying or when credentials are required. ## Quick Start Apply the autonomous progress and escalation rules to continue this multi-step task without unnecessary questions, pausing only when my input or authorization is genuinely required.