subagent-timeout-recovery

What problem does it solve?

This Skill addresses the common pain points of wasted effort when subagents time out mid-task (losing partial completed work) and the risk of delivering unverified subagent self-reports that contain hidden bugs, incomplete changes, or incorrect implementations.

Core Features & Use Cases

• Timeout Recovery Workflow: Step-by-step process to recover partial subagent outputs (e.g., partially built full-stack features) instead of re-delegating the entire task, saving hours of redundant work.
• Trust-but-Verify Validation: Mandatory verification steps for subagent-reported completed work, including typecheck re-runs, file grep checks, and database smoke tests to catch hidden issues before delivery to users.
• Early Stop Signals & Audit Rules: Guidelines to detect subagent typecheck loops before they hit full timeouts, and pre-push audit steps to catch uncommitted partial work from previous sessions that could break production.
• Use Case: For example, if a subagent times out while building a customer relationship management system, you can use this Skill to recover the already created API routes and frontend components, fix any issues, and verify the work without re-running the entire delegation.

Quick Start

Use the subagent-timeout-recovery skill to recover partial work from a timed-out subagent that was building a new full-stack feature, then verify all outputs before delivering them to the user.