From mk
Creates structured multi-file implementation plans before build. Scope-aware: trivial tasks exit early, simple tasks get fast plans, complex tasks get full research + phase files + validation. Enforces Gate 1. Activated by /mk:plan or /mk:cook.
How this skill is triggered — by the user, by Claude, or both
Slash command
/mk:plan-creator [task description] [--fast | --hard | --deep | --parallel | --two | --product-level [--no-design] [--no-scout] | --spike --timebox <duration>] [--tdd] OR [archive | red-team {path} | validate {path}]When to use
Use when creating a structured multi-file implementation plan before build. NOT for ticket analysis (see mk:planning-engine) or scope review (see mk:plan-ceo-review).
[task description] [--fast | --hard | --deep | --parallel | --two | --product-level [--no-design] [--no-scout] | --spike --timebox <duration>] [--tdd] OR [archive | red-team {path} | validate {path}]This skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Scope-aware planning with step-file workflow. Produces plan.md overview + phase-XX detail files.
assets/plan-template.mdassets/product-spec-template.mdassets/spike-plan-template.mdprompts/personas/plan-assumption-destroyer.mdprompts/personas/plan-failure-mode-analyst.mdprompts/personas/plan-scope-complexity-critic.mdprompts/personas/plan-security-adversary.mdreferences/adr-generation.mdreferences/anthropic-example-plan.mdreferences/archive-workflow.mdreferences/cold-start-context-brief.mdreferences/deep-mode.mdreferences/failure-catalog.mdreferences/gate-1-approval.mdreferences/gotchas.mdreferences/ops-metrics-design.mdreferences/output-standards.mdreferences/parallel-mode.mdreferences/phase-template.mdreferences/plan-mutation-protocol.mdScope-aware planning with step-file workflow. Produces plan.md overview + phase-XX detail files.
Activate when:
/mk:plan [task] or /mk:cook [task]--product-level/mk:plan archive → route to references/archive-workflow.md (skip planning pipeline)/mk:plan red-team {path} → route to references/red-team-standalone.md (skip planning pipeline)/mk:plan validate {path} → route to references/validate-standalone.md (skip planning pipeline)Skip when:
/mk:fix with complexity=simple (Gate 1 exception)mk:scale-routing returns workflow=one-shot + zero blast radius| Flag | Mode | Research | Phase Files | Validation |
|---|---|---|---|---|
| (default) | Auto-detect from scope | Follows mode | Follows mode | Follows mode |
--fast | Fast | Skip | plan.md only | Semantic checks only |
--hard | Hard | 2 researchers | plan.md + phases | Full interview |
--deep | Deep | 2-3 researchers | plan.md + phases + bounded phase map: file inventory, test gaps, interfaces, dependencies | Full interview |
--parallel | Parallel | 2 researchers | plan.md + phases + ownership matrix | Full interview |
--two | Two approaches | 2 researchers | 2 approach files + trade-off matrix | After selection |
--product-level | Product spec | 2 researchers (broader) | plan.md only (user stories + features + design language; NO phase files) | Semantic + check-product-spec.sh (no red-team, no validation interview — v1) |
--spike | Time-boxed investigation | Skip | spike plan from assets/spike-plan-template.md (investigate + findings phases; NO test/ship) | Semantic only |
--spike constraints (Agile context only — gated by agile-feedback-cycle.md 2 when loaded):
--timebox <duration> (e.g., --timebox 2d, --timebox 4h); reject with "spike requires --timebox" if absentspike: true, timebox:, and findings_doc: (default tasks/plans/{slug}/findings.md)--product-level and mk:autobuild FULL density (autobuild gate breaks). Reject combination "spike incompatible with autobuild FULL — use --fast or mk:cook"story_points > 5 — likely two spikes, or spike + storyComposable flags:
--tdd — add tests-first phase sections and preserve strict TDD in the cook handoff. See references/tdd-mode.md.Before producing a plan, plan-creator MUST be able to answer all 5 dimensions in concrete sentences (cook's exact-requirements contract — see .claude/skills/cook/SKILL.md "Exact-Requirements Contract (Phase 1)"):
Every clarifying question MUST cite scout findings (file paths) in its options. Abstract options like "Add the feature" without a file path are a failure mode — replace with options of the form "Add to src/api/users.ts (matches existing pattern)" or "Create new src/api/profile.ts".
Skip when input is an existing plan path (plan.md / phase-*.md) — the plan already encodes scout output and the 5 dimensions.
Before starting, read references/failure-catalog.md for common planning failure modes to avoid.
Execute via workflow.md. Step-file architecture — load one step at a time.
Fast mode (--fast) uses workflow-fast.md (steps 00→03→04→07→08).
Agile DoR advisory (Phase 1 entry, conditional): if the parsed plan frontmatter contains a non-empty jira_tickets: list AND agile-story-gates.md is loaded (Agile context active per mk:agent-detector Step 0b), run 1 of agile-story-gates.md for each ticket BEFORE generating phase files. Render the advisory checklist; never block — let the user decide. Skip silently when jira_tickets: is absent or the rule is not loaded.
Step 0: Scope Challenge → trivial (exit) | simple (fast) | complex (hard/deep)
Step 1: Research (hard/deep/parallel/two only) → 2-3 researchers, max 5 calls each
Step 2: Codebase Analysis (hard/deep/parallel/two only) → scout + docs (deep: bounded scope map)
Step 3: Draft Plan → plan.md (≤80 lines) + phase-XX files (12-section template; deep: + phase map; tdd: + regression sections)
Step 4: Semantic Checks → goal/ACs/constraints + structural validation
Step 5: Red Team (hard/deep/parallel/two only) → 4-persona scaling, red-team-findings.md, adjudication
Step 6: Validation Interview (hard/deep/parallel/two only) → 3-5 critical questions with detection keywords, propagate answers
Step 7: Gate 1 → self-check + AskUserQuestion (Approve | Modify | Reject)
Step 8: Hydrate Tasks → phase checkboxes → session tasks
Step 9: Post-Plan Handoff → mode-pruned AskUserQuestion (cook|validate|red-team|end) → write `handoff.next` to plan.md → STOP
Fast mode: Single plan.md (goal, context, scope, approach, ACs)
Hard mode: Directory with plan.md + phase files:
tasks/plans/YYMMDD-name/
├── plan.md ← overview (≤80 lines)
├── research/ ← researcher reports (hard mode)
├── phase-01-name.md ← 12-section detail (≤150 lines)
├── phase-02-name.md
└── ...
--deep to decide architecture: deep maps a chosen approach; if the approach is unclear, route to mk:brainstorming first.tdd_mode=true, step-09 MUST print cook with --tdd; sentinel/env state alone is not a cross-session contract.AskUserQuestion; do NOT auto-invoke the chosen command. User must type it in a fresh session for clean context.handoff.next enum is validated: values outside {cook, validate, red-team, autobuild, end} fail validate-plan.py.references/whole-plan-sweep.md.consistency_sweeps frontmatter is optional: legacy plans without it still validate as PLAN_COMPLETE..plan-state.json v1.2 schema is additive: consumers MUST treat unknown keys (verification_tier, consistency_sweeps_passed) as optional and default-empty. v1.1 readers ignore them silently.references/task-management.md "Post-Hydration Integrity Checks".| File | Purpose |
|---|---|
workflow.md | Step sequence, variable table, flow diagram (hard mode) |
workflow-fast.md | Compact step sequence for --fast mode (00→03→04→07→08) |
step-03a-product-spec.md | Product-level spec drafter: user stories, features, design language. Replaces step-03 when planning_mode = product-level. |
assets/product-spec-template.md | Product spec template (Vision, Features, Design Language, AI Integration, Out-of-Scope) |
assets/spike-plan-template.md | Spike plan template (used when planning_mode = spike). Two phases: investigate + findings. Required frontmatter: spike: true, timebox:, findings_doc:. NOT compatible with mk:autobuild FULL. |
references/anthropic-example-plan.md | RetroForge few-shot calibration example for product-level mode (ambition + feature density reference) |
step-05-red-team.md | Red team review: persona scaling, subagent dispatch, adjudication |
step-06-validation-interview.md | Critical question generation and answer propagation |
step-07-gate.md | Self-check and Gate 1 AskUserQuestion presentation |
step-09-post-plan-handoff.md | Deterministic post-Gate-1 handoff: mode-pruned AskUserQuestion (cook | validate | red-team | autobuild | end), live risk re-scan, writes handoff.next to plan.md frontmatter, prints command + STOPs. |
references/whole-plan-sweep.md | Whole-Plan Consistency Sweep algorithm (Gates W1 + W2). Stage-then-apply: read-only Pass 1 stages Pending Sweep Edits; decision check blocks on unresolved contradictions; write Pass 2 applies edits and logs. |
references/deep-mode.md | Deep mode contract: when to use/avoid, scout budgets, phase-map appendix, context rules. |
references/tdd-mode.md | TDD flag contract: phase sections, optional frontmatter, cook handoff propagation, RED-task hydration. |
references/verification-roles.md | Verification Roles for step-04 sub-step 4d (Fact Checker / Flow Tracer / Scope Auditor / Contract Verifier) with tier selection by phase count. Subagents are READ-ONLY; orchestrator writes the ## Verification Log. |
prompts/personas/plan-assumption-destroyer.md | Plan-specific assumption skeptic persona |
prompts/personas/plan-scope-complexity-critic.md | Plan-specific YAGNI/scope minimalist persona |
prompts/personas/plan-security-adversary.md | Plan-specific security adversary (auth bypass, injection, data exposure). Used at 4+ phases. |
prompts/personas/plan-failure-mode-analyst.md | Plan-specific failure mode analyst (race conditions, cascading failures, recovery). Used at 6+ phases. |
references/phase-template.md | 12-section phase file template |
references/ops-metrics-design.md | If task involves metrics/KPIs/dashboards, load for metric philosophy, templates, and domain fallbacks |
references/cold-start-context-brief.md | When writing phase files, follow this template for self-contained, cold-start-safe phase files |
references/plan-mutation-protocol.md | When modifying an existing plan (split/insert/skip/reorder/abandon), follow this protocol |
references/worked-example-stripe-billing.md | For plan detail level reference, see this complete worked example (the 7-phase model) |
references/scope-challenge.md | Scope modes (HOLD/EXPANSION/REDUCTION) |
references/research-phase.md | Researcher spawning protocol |
references/plan-organization.md | Directory structure, naming, size rules |
references/output-standards.md | YAML frontmatter, required sections |
references/validation-questions.md | Critical question categories for interview |
references/gate-1-approval.md | AskUserQuestion gate + Context Reminder |
references/task-management.md | Hydration, cross-session resume, sync-back |
references/solution-evaluation.md | Trade-off scoring criteria |
references/gotchas.md | Full gotchas list |
references/solution-design-checklist.md | Trade-off analysis checklist for Architecture/Risk/Security sections (5 dimensions) |
references/archive-workflow.md | Archive subcommand: scan completed plans, journal, archive/delete |
references/red-team-standalone.md | Red-team standalone subcommand: adversarial review on existing plan |
references/validate-standalone.md | Validate standalone subcommand: critical question interview on existing plan |
references/adr-generation.md | Architecture Decision Record generation |
references/parallel-mode.md | Ownership matrix template, parallel group rules |
references/two-approach-mode.md | Approach file template, trade-off matrix, selection flow |
scripts/validate-plan.py | Plan completeness validator. Depends on PyYAML (installed via .claude/scripts/bin/setup-workflow into .claude/skills/.venv/). Run via .claude/skills/.venv/bin/python3 scripts/validate-plan.py <plan.md>. |
scripts/check-product-spec.sh | Product-spec structural validator (POSIX bash). Enforces feature count, user stories, forbidden patterns. Used by step-03a and step-04 for --product-level mode. |
references/workflow-models/feature-model.md | Workflow template for feature tasks (loaded JIT by step-00) |
references/workflow-models/bugfix-model.md | Workflow template for bug fix tasks (loaded JIT by step-00) |
references/workflow-models/refactor-model.md | Workflow template for refactor tasks (loaded JIT by step-00) |
references/workflow-models/security-model.md | Workflow template for security tasks (loaded JIT by step-00) |
references/anthropic-example-plan.md — filename retained as the source-provenance attribution for the few-shot calibration example. The file content is the appendix of a published harness-design article (Prithvi Rajasekaran, Labs); renaming would obscure the research origin. Future brand-prose audits skip this filename per this documented exception..claude/rules/gate-rules.md — Gate 1 hard-stop conditions this skill enforces (plan approval before Phase 3).claude/rules/orchestration-rules.md — boundary contract; verification subagents stay READ-ONLY; sweep stays in planner context (never delegated)Read and follow workflow.md.
npx claudepluginhub ngocsangyem/meowkit --plugin mkStructured planning pipeline that discovers, classifies, explores, decomposes, and details medium/complex tasks into a phased plan with cross-cutting concerns and gate checkpoints.
Creates structured plans for multi-step tasks including software features, implementations, research, or projects. Deepens plans via interactive sub-agent reviews.
Generates structured implementation plans from feature descriptions or requirements, grounded in repo patterns and research. Deepens existing plans via interactive sub-agent review.