What problem does it solve? Turning a vague goal, scattered notes, or an existing spec into something an agent can safely execute usually requires separate brainstorming and planning steps, and unspoken assumptions slip through. This Skill replaces that with one interactive dialogue that surfaces every load-bearing decision and produces a validated three-document contract before any code is touched. ## Core Features & Use Cases - Interactive intent elicitation: Decomposes any input (goal, docs, notes, or a mix) into challengeable material, then extracts domain decisions from the user and presents technical options with trade-offs instead of guessing. - Independent audits: Dispatches two fresh subagent checks — an intent-completeness guess-hunt before the spec is frozen, and a final 3-doc gate — so un-grounded assumptions are caught while the user can still decide. - Execution-ready 3-doc output: Writes handoff.md, spec.md, and plan.md (with a dependency graph) into .dryforge/, ready for the go skill to execute. - Use Case: You have a rough feature idea and a brain-dump from another tool. Invoke the ready skill, answer its targeted questions, review the generated spec and plan, then hand off to go for execution. ## Quick Start Invoke the ready skill with a description of what you want to build or change, then answer its questions and approve the resulting three-document plan.