菜单
Decompose a core problem into root causes with an ASCII tree. Outputs a fenced code block using box-drawing characters. Use for root-cause analysis, drilling from a high-level problem down to testable sub-problems, or step 2 of a strategic analysis.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | issue-trees |
| description | Decompose a core problem into root causes with an ASCII tree. Outputs a fenced code block using box-drawing characters. Use for root-cause analysis, drilling from a high-level problem down to testable sub-problems, or step 2 of a strategic analysis. |
An issue tree breaks a vague problem into progressively smaller, more concrete sub-problems until you reach something testable. Each branch is a hypothesis about why the parent problem exists. Done well, the leaves are things you can actually go investigate.
A fenced code block with text syntax. Use ASCII box-drawing characters:
├── non-final child│ vertical continuation under a non-final child└── final child at each level```text
Core Problem
├── Category 1
│ ├── Sub-factor A
│ └── Sub-factor B
└── Category 2
└── Sub-factor C
```
Some destinations mangle box-drawing characters — Slack, certain email clients, and some chat UIs render them as boxes or question marks. If you know the output is heading to one of those, use an indented-dash tree instead:
- Core Problem
- Category 1
- Sub-factor A
- Sub-factor B
- Category 2
- Sub-factor C
It loses some elegance but stays readable everywhere. Default to box-drawing; switch only when needed.
This is "Slack mode." When the user asks for Slack / Teams / email / plain-text output (or names that as the destination), use this indented-dash form for the whole tree by default.
| Default | When to flex |
|---|---|
| At least 2 levels deep | A 1-level tree is a list. Push deeper. For complex problems, 3–4 levels. |
| 3–8 words per node | Stay short — long node text destroys readability. |
| Branches MECE at each level | This one doesn't flex. |
| Stop when leaves are testable | A leaf that says "we have problems with customers" isn't done. "Onboarding completion fell to 61%" is. |
Problem: E-commerce site conversion rate dropped from 3.2% to 1.9% over Q3.
Conversion 3.2% → 1.9%
├── Traffic mix shift
│ ├── Paid social spend redirected to top-of-funnel
│ ├── Branded search share fell
│ └── Affiliate program reduction
├── On-site experience
│ ├── Page-load time +800ms after CDN change
│ ├── New checkout layout shipped Aug 14
│ └── Mobile bounce rate climbing
├── Pricing & promotion
│ ├── Promo cadence reduced (cost-cutting)
│ ├── Free-shipping threshold raised $50 → $75
│ └── Competitor undercutting on top 20 SKUs
└── Inventory & merchandising
├── 12 hero SKUs out of stock
├── New collection launch delayed two weeks
└── Recommendation algorithm retrained
Every leaf is concrete. Each one could be assigned to someone with "validate this in 3 days." That's the bar.
strategy-consultant — issue trees are step 2 of 5. If you also want to test which branch is the actual cause, follow up with hypothesis-driven.Adapted from Analyst Academy on YouTube — 5 Consulting Frameworks to Solve Any Problem. MIT-licensed; see LICENSE.
Apply the 5 Whys technique to drill from a surface problem to a root cause. Outputs five sequentially numbered "Why?" lines with brief answers, ending in a root-cause statement. Use after an issue tree to drill the critical branch deeper, or any time the user wants to test whether the apparent cause is the actual cause.
Structure a recommendation top-down — Barbara Minto's Pyramid Principle. Outputs a single-sentence top-line answer, three to five supporting arguments, and one layer of evidence under each. Use to convert analysis into executive communication, or to structure a board-meeting talk track.
Tier-1 strategy-consultant analysis using seven frameworks in sequence — MECE, Issue Trees, 5 Whys, Hypothesis-Driven, Pareto, So What?, and Pyramid Principle. Use when the analysis needs both deeper root-cause drilling AND executive-ready communication structure. Heavier than the 5-framework master; reach for it on high-stakes work.
Tier-1 strategy-consultant analysis tailored for edtech / online-learning problems — activation, completion, learning outcomes, monetization. Same five frameworks with edtech-aware MECE defaults, vocabulary, and root-cause priors.
Tier-1 strategy-consultant analysis tailored for fintech / payments / lending / fraud problems — approval rates, fraud loss, delinquency, regulatory friction. Same five frameworks with fintech-aware MECE defaults, vocabulary, and root-cause priors.
Tier-1 strategy-consultant analysis tailored for public-sector / civic-ops problems — backlogs, processing cycle time, citizen satisfaction, procurement. Same five frameworks with government-aware MECE defaults, vocabulary, and root-cause priors.