What problem does it solve? Agents drift between roles, forget decisions, and start editing code before the goal is clear. This Skill enforces a routing-first workflow so every repo task passes through clarification, intake, execution, verification, and closeout instead of ad-hoc prompting. ## Core Features & Use Cases - Mandatory skill routing: loads first for repo work and directs each turn to the correct workflow skill (clarification, intake, execution, verification, or lesson capture). - Clarification and anti-sycophancy gates: blocks edits and subagent dispatch until the expected effect and plan are approved, and challenges weak user premises before optimizing. - Skill-only fallback: when runtime plugins are unavailable, it self-enforces task readiness via just-demand . list-active before any write or subagent dispatch. - Use Case: A user reports "the export button stopped working." Instead of patching immediately, the agent loads this skill, runs socratic clarification to pin down the expected behavior, creates an intake, promotes a formal task, then executes and verifies through the workflow. ## Quick Start Ask the agent to handle a bug report or feature request in this repo and it will load this skill first to route the work through clarification and the formal task workflow.
