메뉴
JTBD workflow classification and routing - ODI two-phase framework, five job types with workflow sequences, baseline type selection, workflow anti-patterns, and common recipes
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
Cross-agent collaboration protocols, workflow handoff patterns, and commit message formats for TDD/Mikado/refactoring workflows
Orchestrates the full DELIVER wave end-to-end (roadmap > execute-all > finalize). Use when all prior waves are complete and the feature is ready for implementation.
Acceptance test creation methodology for the DISTILL wave. Domain knowledge for the acceptance designer agent: port-to-port principle, prior wave reading, wave-decision reconciliation, graceful degradation, and document back-propagation.
Creates a phased roadmap.json for a feature goal with acceptance criteria and TDD steps. Use when planning implementation steps before execution.
Dispatches one unit of DELIVER work to a specialized agent for TDD execution. Runs a single roadmap.json step through the TDD cycle.
Deep knowledge for Outside-In TDD - double-loop architecture, ATDD integration, port-to-port testing, walking skeletons, and test doubles policy
| name | nw-jtbd-workflow-selection |
| description | JTBD workflow classification and routing - ODI two-phase framework, five job types with workflow sequences, baseline type selection, workflow anti-patterns, and common recipes |
| user-invocable | false |
| disable-model-invocation | true |
Classify incoming work by job type and recommend the appropriate nWave workflow entry point. Use during Phase 1 (GATHER) to triage before crafting stories.
Determine which phase applies before proceeding.
Phase 1: Discovery -- when you do not know what to build
[research] --> discuss --> design --> distill
| | | |
GATHER WHAT are HOW should WHAT does
evidence the needs? it work? "done" look like?
Phase 2: Execution Loop -- when you know what needs to change
[research] --> baseline --> roadmap --> split --> execute --> review
| | | | | |
GATHER MEASURE PLAN it BREAK it DO each CHECK
evidence first completely into atoms task quality
|
<-----------+ (loop per task)
Key insight: research is a cross-wave capability invocable at any point for evidence-based decisions.
Skip discovery and enter execution loop directly when ALL hold:
If any fail, start with discovery (DISCUSS wave).
"I need to create something that doesn't exist yet"
[research] -> discuss -> design -> [diagram] -> distill -> baseline -> roadmap -> split -> execute -> review
| Step | Purpose |
|---|---|
| research | (Optional) Gather domain knowledge before requirements |
| discuss | Gather requirements -- you don't know what's needed yet |
| design | Architecture decisions, technology selection |
| diagram | (Optional) Visualize architecture for stakeholders |
| distill | Define acceptance tests -- what does "done" look like? |
| baseline | Measure starting point for tracking improvement |
| roadmap | Comprehensive plan while context is fresh |
| split | Break into atomic, self-contained tasks |
| execute | Do each task with clean context |
| review | Quality gate before proceeding |
"I know what needs to change in our system"
[research] -> baseline -> roadmap -> split -> execute -> review (repeat)
Skip discovery: system understood and problem identified. Baseline is blocking gate -- measure current state before planning. Prevents "optimizing the wrong thing."
"Code works but structure needs improvement"
Simple refactoring:
[root-why] -> mikado -> refactor (incremental)
Complex refactoring with tracking:
[research] -> baseline -> roadmap (methodology: mikado) -> split -> execute -> review
Mikado Method explores dependencies before committing. Reversible at every step.
"Something is broken and I need to find why"
[research] -> root-why -> develop -> deliver
Minimal sequence -- focused intervention only.
"I need to gather information before deciding"
research -> [decision point: which job to pursue next]
No execution -- pure information gathering feeding into other jobs.
| Job | You Know What? | Sequence |
|---|---|---|
| Greenfield | No | [research] -> discuss -> design -> [diagram] -> distill -> baseline -> roadmap -> split -> execute -> review |
| Brownfield | Yes | [research] -> baseline -> roadmap -> split -> execute -> review |
| Refactoring | Partially | [research] -> baseline -> mikado/roadmap -> split -> execute -> review |
| Bug Fix | Yes (symptom) | [research] -> root-why -> develop -> deliver |
| Research | No | research -> (output informs next job) |
Items in [brackets] are optional. Cross-wave commands (usable anytime): research, diagram, root-why, git.
When workflow includes a baseline step, advise on which type to create.
Use when improving speed, reducing resource usage, or optimizing throughput. Required: timing measurements with breakdown | bottleneck ranking | target metrics with evidence | quick wins identified.
Use when fixing workflow issues, preventing incidents, or improving reliability. Required: incident references or failure modes | simplest alternatives considered (with why insufficient).
Use when building new capabilities (greenfield or brownfield). Required: current state analysis | requirements source and validation.
Operate at project/feature level, distinct from story-level anti-patterns in leanux-methodology skill.
| Anti-Pattern | Problem | Solution |
|---|---|---|
| Skip research | Decisions without evidence | Research when unfamiliar with domain |
| Skip baseline | Optimize the wrong thing | Always baseline before roadmap |
| Monolithic tasks | Context degradation | Use split for atomic tasks |
| Skip review | Quality issues propagate | Review before each execute |
| Architecture before measurement | Over-engineering | Baseline identifies quick wins first |
| Forward references in tasks | Tasks not self-contained | Each task must have all context embedded |
| Situation | Entry Point | Key Characteristic |
|---|---|---|
| New feature on existing codebase | baseline (skip discovery) | Existing system, new capability |
| Performance optimization | baseline (type: performance) | Measurement-first |
| Legacy system modernization | research + root-why + baseline | Deep understanding first |
| Quick bug fix | root-why + develop + deliver | Minimal sequence |
| Pure research task | research | Output informs next job selection |
| Data-heavy project | research + baseline | Specialist agent involvement |
| Category | Core Job |
|---|---|
| Understanding | Know what to build and why |
| Planning | Break work into safe, trackable chunks |
| Executing | Do work without context degradation |
| Validating | Catch issues early with quality gates |
| Communicating | Share understanding via diagrams and docs |
| Investigating | Find truth before acting |
For deep opportunity analysis with ODI scoring, defer to product-discoverer agent. Product-owner applies simpler prioritization (MoSCoW, Value/Effort) for story-level ordering -- see leanux-methodology skill.