Decompose

Learn to decompose complex problems before building real agents. Better answers today. Ready for orchestration tomorrow. Breaks messy tasks into clear DAG structures with explicit dependencies, parallel branches, and synthesis logic. Improves Claude outputs by enforcing topological order and catching contradictions. Teaches the mental model for LangGraph, CrewAI, n8n, and Sub-Agents. Type "decompose: [goal]" or "decompose preview: [goal]" to start. Zero experience required.

15 stars 1 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add matt-k-wong/mkw-DAG-architect --skill decompose

Manual install

git clone --depth 1 https://github.com/matt-k-wong/mkw-DAG-architect.git /tmp/mkw-DAG-architect
mkdir -p ~/.claude/skills
cp -r /tmp/mkw-DAG-architect/. ~/.claude/skills/decompose

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

You are now in decompose mode. Instead of streaming a single linear response, you break
down complex goals into a structured Directed Acyclic Graph of dependent tasks, execute
them in topological order, simulate parallel branches, and synthesize a final output that
is explicitly better than what vanilla linear reasoning produces.

---

## IMPACT PREVIEW MODE  ← The "Do I Need This?" Feature

**Trigger:** User types `decompose preview: [goal]` OR `decompose: [goal]` OR `Should I decompose [goal]?`

When triggered, respond ONLY with this block — do NOT execute the full DAG:

```
╔══════════════════════════════════════════════╗
║           DAG IMPACT PREVIEW                 ║
╚══════════════════════════════════════════════╝
Goal: [restate goal concisely]

Verdict: ✅ HIGH VALUE  /  ⚠️ MARGINAL  /  ❌ SKIP

Without DAG (vanilla Claude):
  ✗ [Specific weakness #1 — dependency it would ignore]
  ✗ [Specific weakness #2 — contradiction it would risk]
  ✗ [Specific weakness #3 — branch it would skip or flatten]

With DAG Simulator:
  ✓ [What topological ordering unlocks]
  ✓ [What parallel branching explores]
  ✓ [What synthesis aggregates that linear misses]

Estimated structure:
  Nodes: ~N  |  Depth: ~D  |  Parallel sets: ~P

  [1] Foundation ──► [2] Branch A ──► [4] Synthesis
                 └──► [3] Branch B ──►

Type "decompose go" or "proceed" to run the full decomposition.
```

**Verdict criteria:**

Read the full skill on GitHub

Details

Skill name: decompose
Repository: matt-k-wong/mkw-DAG-architect
License: MIT
Repo updated: 15 March 2026

FAQ

What does the Decompose skill do?

How do I install the Decompose skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add matt-k-wong/mkw-DAG-architect --skill decompose. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Decompose popular?

It lives in matt-k-wong/mkw-DAG-architect, which has 15 GitHub stars and 1 forks.

Related skills

Format Agent

KaguraNanaga/format-agent-skill

按规范文字、内置 Style Pack 或 Word 模板重排以段落和常规表格为主的中文/英文 Word 文稿。适用于公文、论文、普通报告、基础技术手册和法律 brief；可有损导入 .doc/.wps/.odt/.rtf。不要用于简历、表单、宣传册、Newsletter、海报、PDF 版面还原或内容控件应用。

AI, LLMs & Agents

Claude Managed Agents

0xArx/claude-managed-agents-skill

Build applications with the Claude Managed Agents API — Anthropic's managed infrastructure for running Claude as an autonomous agent in a cloud container. Trigger this skill whenever the user wants to build with managed agents, asks about agent sessions, wants to stream agent events, uses client.beta.agents / client.beta.sessions / client.beta.environments, asks about the managed-agents-2026-04-01 beta, wants Claude to autonomously run tools in a sandbox, or asks about outcomes, multiagent orchestration, or persistent agent memory. Also trigger when the user asks how to run Claude for long-run

AI, LLMs & Agents

Multi LLM

beastlabai/multi-llm-plugin

Orchestrate code plan reviews, task generation, implementation, code reviews and commit reviews using multiple LLM providers in parallel, or ask each model a free-text question about a plan. Use this skill whenever the user wants to review a plan with multiple models, generate implementation tasks, implement tasks with subagent delegation, review code changes against a plan, review or fix the commits they made, ask the models a question about a plan, or run any multi-model workflow. Triggers on: 'review my plan', 'multi-llm', 'run code review', 'generate tasks from plan', 'implement this plan'

AI, LLMs & AgentsIn 1 list

Chainaware Behavioral Prediction

ChainAware/behavioral-prediction-mcp

Use this skill whenever a user asks about wallet safety, fraud risk, rug pull detection, wallet behavior analysis, DeFi personalization, on-chain reputation scoring, AML checks, token ranking by holder quality, airdrop screening, lending risk, token launch auditing, or AI agent trust scoring. Triggers on questions like: is this wallet safe?, will this pool rug pull?, what will this address do next?, score this wallet, detect fraud for address, personalize my DeFi agent, rank this token, top AI tokens, best holders of this token, check this contract, is this token safe?, profile this wallet, KY

AI, LLMs & AgentsIn 1 list

Cavil Loop

GigleAI/cavil-loop

GitHub label-driven workflow that lets a local AI handle issues / PRs; the daemon listens for `pending/agent` and dispatches work. Every issue's design proposal / discussion / code / Claude conversation / tmux history is filed under the issue number, so you can find or resume any of it later just by knowing the issue number. Commands include setup (deploy to a project), status (check daemon state), disable (turn off a project's daemon)

AI, LLMs & Agents

Bug Hunting

m4vic/bug-hunting

Bug bounty hunting and penetration testing skills for Claude, Codex, and other agentic AI tools.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.