// [Planning] Use when you need to start a new project from scratch with full waterfall inception — idea, research, domain modeling, tech stack, and implementation plan.
[HINT] Download the complete skill directory including SKILL.md and all related files
| name | greenfield |
| version | 1.1.0 |
| description | [Planning] Use when you need to start a new project from scratch with full waterfall inception — idea, research, domain modeling, tech stack, and implementation plan. |
Goal: Guide greenfield project inception from raw idea to an approved, implementable project plan using a full waterfall process.
Workflow (16 steps):
/idea) — Interview user about problem, vision, constraints, team skills, scale. DO NOT ask about tech stack — keep business-focused./web-research) — WebSearch for competitors, market landscape, existing solutions/deep-research) — WebFetch top sources, extract key findings/business-evaluation) — Viability assessment, risk matrix, value proposition/domain-analysis) — Bounded contexts, aggregates, entities, ERD diagram, domain events. Validate every context boundary with user./tech-stack-research) — Derive technical requirements from business + domain analysis. Research top 3 options per stack layer (backend, frontend, database, messaging, infra). Detailed pros/cons matrix, team-fit scoring, market analysis. Present comparison report for user to decide./architecture-design) — Research and compare top 3 architecture styles (Clean, Hexagonal, Vertical Slice, etc.). Evaluate design patterns (CQRS, Repository, Mediator). Audit against SOLID, DRY, KISS, YAGNI. Validate scalability, maintainability, IoC, technical agnosticism. Present comparison with recommendation. Harness output required: produce a "Scaffold Handoff — Harness Plan" table in the architecture report: (a) feedforward guides to create (AGENTS.md sections, skill activation rules, pattern catalog), (b) computational feedback sensors to install (linter, formatter, pre-commit, CI), (c) inferential feedback sensors to configure (review skills, AI gates). This table feeds /scaffold → /linter-setup → /harness-setup./plan) — Create phased plan using confirmed tech stack + architecture + domain model/security) — Review plan for OWASP Top 10, auth patterns, data protection concerns/performance) — Review plan for performance bottlenecks, scalability, query optimization/plan-review) — Full plan review, risk assessment, approval/refine) — Transform idea + reviewed plan into actionable PBI with acceptance criteria/story) — Break PBI into implementable user stories/plan-validate) — Interview user with critical questions to validate plan + stories/tdd-spec) — Test pyramid, frameworks, spec outline/workflow-end) — Clean up, announce completionKey Rules:
AskUserQuestion validation before proceedingsolution-architect agent/linter-setup handles install), and inferential sensors (/harness-setup configures). The scaffold + linter-setup + harness-setup triad are NON-SKIPPABLE infrastructure — code without a harness accumulates technical debt from day one.This skill is the explicit entry point for the greenfield-init workflow.
When invoked:
greenfield-init workflow via /workflow-start greenfield-initsolution-architect agent provides architecture guidance throughout/plan or /feature instead)/cook-fast instead)All artifacts saved to plan directory:
plans/{id}/
research/
discovery-interview.md
market-research.md
deep-research.md
business-evaluation.md
domain-analysis.md
tech-stack-comparison.md
architecture-design.md
phase-01-domain-model.md
phase-02-tech-stack.md
phase-02b-architecture.md
phase-03-project-structure.md
phase-04-test-strategy.md
phase-05-backlog.md
plan.md (master plan with YAML frontmatter)
After completion, recommend next step: /cook to scaffold the project structure.
Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).
MANDATORY IMPORTANT MUST ATTENTION break work into small todo tasks — one per workflow step. MANDATORY IMPORTANT MUST ATTENTION validate with user at EVERY step — never auto-decide. MANDATORY IMPORTANT MUST ATTENTION add a final review todo task to verify work quality and identify fixes/enhancements.
MANDATORY IMPORTANT MUST ATTENTION use TaskCreate to break ALL work into small tasks BEFORE starting.
MANDATORY IMPORTANT MUST ATTENTION use AskUserQuestion at EVERY stage — validate decisions before proceeding.
MANDATORY IMPORTANT MUST ATTENTION NEVER ask tech stack upfront — business analysis and domain modeling first.
AI Mistake Prevention — Failure modes to avoid on every task:
Check downstream references before deleting. Deleting components causes documentation and code staleness cascades. Map all referencing files before removal. Verify AI-generated content against actual code. AI hallucinates APIs, class names, and method signatures. Always grep to confirm existence before documenting or referencing. Trace full dependency chain after edits. Changing a definition misses downstream variables and consumers derived from it. Always trace the full chain. Trace ALL code paths when verifying correctness. Confirming code exists is not confirming it executes. Always trace early exits, error branches, and conditional skips — not just happy path. When debugging, ask "whose responsibility?" before fixing. Trace whether bug is in caller (wrong data) or callee (wrong handling). Fix at responsible layer — never patch symptom site. Assume existing values are intentional — ask WHY before changing. Before changing any constant, limit, flag, or pattern: read comments, check git blame, examine surrounding code. Verify ALL affected outputs, not just the first. Changes touching multiple stacks require verifying EVERY output. One green check is not all green checks. Holistic-first debugging — resist nearest-attention trap. When investigating any failure, list EVERY precondition first (config, env vars, DB names, endpoints, DI registrations, data preconditions), then verify each against evidence before forming any code-layer hypothesis. Surgical changes — apply the diff test. Bug fix: every changed line must trace directly to the bug. Don't restyle or improve adjacent code. Enhancement task: implement improvements AND announce them explicitly. Surface ambiguity before coding — don't pick silently. If request has multiple interpretations, present each with effort estimate and ask. Never assume all-records, file-based, or more complex path.
Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.
MUST ATTENTION apply critical thinking — every claim needs traced proof, confidence >80% to act. Anti-hallucination: never present guess as fact.
MUST ATTENTION apply AI mistake prevention — holistic-first debugging, fix at responsible layer, surface ambiguity before coding, re-read files after compaction.
TaskCreate BEFORE startingfile:line evidence for every claim (confidence >80% to act)[TASK-PLANNING] Before acting, analyze task scope and systematically break it into small todo tasks and sub-tasks using TaskCreate.