Conversational wizard that collects a plain-language feature description, derives a normalized feature ID, checks for naming conflicts, asks clarifying questions, classifies the feature (backend, frontend, infra, cross-cutting), and recommends the next nWave wave to run (discover, discuss, design, distill, deliver). Once confirmed, it can create the feature directory and invoke the chosen wave command to bootstrap work.
Use this skill when you have an idea or request but aren’t sure where to start in the nWave pipeline: new features, bug-fix paths, or when you need to validate whether discovery or discussion is required. Triggers include command-style inputs like /nw-new "Add rate limiting to the API gateway", brief feature statements, or when a user asks "where do I start with X?"
Best suited to agentic frameworks that support CLI-style commands and subagent orchestration (Claude Code / Claude-based workflows, agentic CLIs).
This skill has not been reviewed by our automated audit pipeline yet.
Diverger Review Criteria
Checks rigor and quality of DIVERGE-wave artifacts: JTBD extraction, research coverage, option diversity, taste scoring, and recommendation traceability.
nWave Agent Creation Workflow
A practical five-phase workflow for designing, building, validating, and refining agent definitions and skills.
Outcome KPI Framework
A practical framework for writing measurable outcome-focused KPIs for features and user stories, mapping outcomes to leading indicators and measurement plans.