How do I install the Breakdown Problem skill?

Run npx skillmds@latest add extremeclarity/breakdown-problem in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Breakdown Problem skill do?

Breaks problems and goals into independent parts that can be tackled separately. It is listed under Coding & Dev Tools on SkillMD.

Is Breakdown Problem safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Breakdown Problem?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Breakdown Problem free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Breakdown Problem?

extremeclarity (@extremeclarity) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Breakdown Problem by extremeclarity

Name: Breakdown Problem
Rating: 4.2 (5 reviews)
Author: extremeclarity

File contents

Introduction

Goal

Break a complex problem or goal into sub-parts for clearer understanding. Focus on what the problem IS, not how to solve it.

Method

Propose a breakdown, then iteratively validate and refine using a context-blind agent until success criteria are met. Expect multiple passes.

Success Criteria

MECE — Sub-parts are mutually exclusive and collectively exhaustive
Uniformity — Sub-parts are at the same level of abstraction
Precision — Each sub-part is well-defined
Problem-focused — Sub-parts describe problems or goals, not solutions
Actionable — Each sub-part can be worked on independently

Workflow

Across all steps, focus on what the problem or goal IS, not how to solve it.

1. Propose breakdown

Think about the problem or goal the user has shared. Use conversation history for relevant details and clues. Do not look for additional context in files or memory.
Try standard decomposition archetypes before settling. Break by lifecycle stage, by stakeholder, by component, or by driver tree (the quantities that combine into the whole). Pick the cut that fits the problem best.
Propose a breakdown into sub-parts. Self-critique against the success criteria. Describe what you observe, not why it happens — stay far from causal or solution framing.

2. Validate

Spawn a blind-reasoner agent (Sonnet). Name it: "Validate breakdown". Provide the agent with the original problem/goal and your proposed breakdown. Ask the agent to validate against the success criteria and surface any gaps or inconsistencies.

3. Revise

If the validator surfaced issues, revise the breakdown. Make the smallest changes that address the issues. Preserve the structure of the original breakdown. Do NOT rewrite from scratch.
Send the revised breakdown back to the validator (step 2). Repeat until the validator passes all success criteria.

Final output

Present breakdown: Present the sub-parts to the user naturally in conversation. For each sub-part, provide a brief description of what it covers.
Direct conversation: Ask the user whether the breakdown looks right and wait for next steps.

Related skills

clarify-framing — If the problem or goal is ambiguous, run clarify-framing first to clarify before decomposing.

---name: breakdown-problemdescription: Breaks problems and goals into independent parts that can be tackled separately.---# Introduction## GoalBreak a complex problem or goal into sub-parts for clearer understanding. Focus on what the problem IS, not how to solve it.## MethodPropose a breakdown, then iteratively validate and refine using a context-blind agent until success criteria are met. Expect multiple passes.### Success Criteria- **MECE** — Sub-parts are mutually exclusive and collectively exhaustive- **Uniformity** — Sub-parts are at the same level of abstraction- **Precision** — Each sub-part is well-defined- **Problem-focused** — Sub-parts describe problems or goals, not solutions- **Actionable** — Each sub-part can be worked on independently## Workflow- Across all steps, focus on what the problem or goal IS, not how to solve it.### 1. Propose breakdown- Think about the problem or goal the user has shared. Use conversation history for relevant details and clues. Do not look for additional context in files or memory.- Try standard decomposition archetypes before settling. Break by lifecycle stage, by stakeholder, by component, or by driver tree (the quantities that combine into the whole). Pick the cut that fits the problem best.- Propose a breakdown into sub-parts. Self-critique against the success criteria. Describe what you observe, not why it happens — stay far from causal or solution framing.### 2. Validate- Spawn a blind-reasoner agent (Sonnet). Name it: "Validate breakdown". Provide the agent with the original problem/goal and your proposed breakdown. Ask the agent to validate against the success criteria and surface any gaps or inconsistencies.### 3. Revise- If the validator surfaced issues, revise the breakdown. Make the smallest changes that address the issues. Preserve the structure of the original breakdown. Do NOT rewrite from scratch.- Send the revised breakdown back to the validator (step 2). Repeat until the validator passes all success criteria.## Final output- Present breakdown: Present the sub-parts to the user naturally in conversation. For each sub-part, provide a brief description of what it covers.- Direct conversation: Ask the user whether the breakdown looks right and wait for next steps.## Related skills- **clarify-framing** — If the problem or goal is ambiguous, run clarify-framing first to clarify before decomposing.

Frequently asked questions