Nw Mikado Method logo

Nw Mikado Method

Organization
nWave-ai
nw-mikado-method

Enhanced Mikado Method for complex architectural refactoring - systematic dependency discovery, tree-based planning, and bottom-up execution

Overview

PublishernWave-ai
RepositorynWave
Skill namenw-mikado-method
Stars
616
Forks
65
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by nWave-ai on GitHub. Read the source before you install it.

Installation

Install the Nw Mikado Method AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/nWave-ai/nWave/tree/main/nWave/skills/nw-mikado-method
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/nWave-ai/nWave.git /tmp/nWave
mkdir -p .claude/skills
cp -r /tmp/nWave/nWave/skills/nw-mikado-method .claude/skills/nw-mikado-method
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Nw Mikado Method in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Nw Mikado Method on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Nw Mikado Method is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Mikado Method

Use for complex refactoring where direct implementation causes cascading failures across multiple classes/modules.

When to Use

  • Refactoring goal affects multiple classes/modules | Direct implementation causes cascade of failures
  • Dependencies not immediately clear | High risk of breaking existing functionality

Core Process

Cycle: Set Goal > Experiment > Visualize prerequisites > Revert to working state.

Treat compilation/test failures as valuable information -- each failure reveals a prerequisite node in the dependency graph. Revert keeps codebase shippable at all times.

Discovery-Tracking Protocol

Commit immediately after each dependency discovery to preserve exploration history and enable interrupt/resume.

Commit Formats

  • Dependency: Discovery: [Class.Method(params)] requires [Prerequisite] in [File:Line]
  • False leaf: Discovery: False leaf - [Node] blocked by [Dependency]
  • Exploration complete: Discovery: No new dependencies found - exploration complete for [GoalArea]
  • Ready: Ready: True leaves identified - [Count] leaves ready for execution

Exhaustive Exploration Algorithm

Sequence: EXPERIMENT > LEARN > GRAPH > COMMIT GRAPH > REVERT

  1. Experiment: Attempt naive implementation of stated goal
  2. Learn: Capture ALL compilation and test failures immediately
  3. Graph: Create concrete prerequisite nodes with exact specifications
  4. Commit Graph: Commit discovery with mandatory format
  5. Revert: Revert ALL code changes to maintain clean state

Termination Criteria

  • Every apparent leaf candidate systematically attempted | No new dependencies emerge from leaf attempts
  • Tree structure stable across multiple exploration cycles | True leaves confirmed with zero prerequisites

Concrete Node Specification

Nodes require method-level specificity:

  • Method signatures: ClassName.MethodName(parameter types) -> ReturnType
  • File locations: src/Services/UserService.cs, line 45
  • Access modifiers: public | private | internal | protected
  • Refactoring technique: Extract Method | Move Method | etc.
  • Atomic transformation: Rename | Extract | Inline | Move | Safe Delete
  • Code smell target: Long Method | Feature Envy | etc.

Tree File Management

File Structure

  • Directory: docs/mikado/ | Filename: <goal-name>.mikado.md
  • Format: - [ ] pending, - [x] completed | Indentation: 4 spaces per nesting level
  • Dependencies indented deeper than dependents

Tree Structure Rules

  1. Root goal at 0 indentation
  2. Direct dependencies at 4-space indentation
  3. Sub-dependencies at 8-space, continuing per level
  4. Child nodes must complete before parent nodes
  5. Nodes at same indentation level are independent (parallelizable)

Two-Mode Operation

Exploration Mode

  1. Attempt naive implementation of refactoring goal
  2. Capture compilation/test failures with full details
  3. Create concrete prerequisite nodes with method-level specificity
  4. Add dependencies to tree file with proper indentation nesting
  5. Commit tree discovery only (specific format)
  6. Revert code changes completely except tree file
  7. Repeat until no new dependencies discovered

Execution Mode

  1. Identify deepest indentation level with incomplete nodes
  2. Select only true leaves (most nested, zero prerequisites)
  3. Execute one leaf at a time for safety
  4. Complete all nodes at current level before moving up
  5. Never attempt parent node until all children complete
  6. Implement minimal possible change per leaf
  7. Validate with full test execution
  8. Commit implementation, update tree marking node complete
  9. Proceed bottom-up to next true leaf

Tree Example

markdown
- [ ] Goal: Replace direct DB calls in OrderController with repository pattern
    - [ ] Update OrderController constructor to use IOrderRepository
        - [ ] Implement SqlOrderRepository : IOrderRepository
            - [ ] Create IOrderRepository interface
                - [ ] Define GetOrderById(int orderId) -> Order? method signature
                - [ ] Define SaveOrder(Order order) -> Task method signature
            - [ ] Add constructor SqlOrderRepository(IDbContext context)
                - [ ] Verify IDbContext is registered in DI container
        - [ ] Implement GetOrderById method
            - [ ] Handle null order case with OrderNotFoundException
                - [x] Create OrderNotFoundException class
    - [ ] Register IOrderRepository in DI container
    - [ ] Remove IDbContext _context field from OrderController

Execution order: deepest leaves first, working up level by level.

Timeboxed Experimentation

10-minute timebox per attempt. If change can't be made in 10 min, it's too complex -- break down further.

  • Success: commit, check off node, move to next
  • Fail: revert, identify missing prerequisites, write subgoals

Goal Definition

Convert technical goals to stakeholder-understandable business value:

  • Correct: "Customer address is retrieved using the latest version of the third-party API for improved reliability"
  • Incorrect: "Update third-party API to version X"

Integration with TDD

  • All Mikado execution steps maintain passing tests (green bar)
  • Each leaf execution = one atomic commit with tests passing
  • If any step breaks tests, revert immediately and reassess dependencies
  • Baby steps rhythm: test-commit-integrate every small change

Frequently asked questions

What does the Nw Mikado Method AI skill do?

Enhanced Mikado Method for complex architectural refactoring - systematic dependency discovery, tree-based planning, and bottom-up execution

Why use Nw Mikado Method on TypingMind?

Because you install it once and use it with any model. Nw Mikado Method is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Nw Mikado Method in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/nWave-ai/nWave/tree/main/nWave/skills/nw-mikado-method. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Nw Mikado Method?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Nw Mikado Method?

As many as you like. As long as a model supports skills, you can use Nw Mikado Method with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Nw Mikado Method AI skill free?

Yes. It is published on GitHub by nWave-ai under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇