What problem does it solve? Turning a written plan or conversation into a runnable Invoker workflow requires strict YAML structure, review-compression headings, layered decomposition metadata, and deterministic validation — doing this by hand produces plans that fail submission gates or silently drop work. ## Core Features & Use Cases - Plan-to-YAML conversion: Transforms Markdown plans, conversations, or policy documents into Invoker YAML plans with required top-level fields (name, onFinish, mergeMode, repoUrl or scratch, tasks). - Deterministic validation gate: Runs skill-doctor.sh plus completeness, atomicity, coverage-map, and stack-manifest checks so plans pass schema, zero-context prompt, and review-unit requirements before submission. - Stack-first and formula-first authoring: Splits multi-slice implementations into chained workflow YAML files and renders proven formula templates (bugfix, entity-research) instead of hand-authoring shapes. - Use Case: Given a feature plan for a multi-package refactor, generate a stack of step-N workflow YAML files, validate them with skill-doctor, and submit the head workflow for review before fan-out. ## Quick Start Ask the agent to convert your plan into an Invoker plan, for example: convert this plan to invoker and validate it before submission.