Cline

Cline Skills

Custom instruction sets for Cline that unlock specialized workflows. Copy and paste into Cline's Custom Instructions field.

How to use

  1. 1. Open Cline in VS Code and click the Settings gear icon.
  2. 2. Scroll to Custom Instructions.
  3. 3. Paste any skill below into the field.
  4. 4. Done — Cline will follow the skill on every task.

MegaBrain Design System

Instructs Cline to create HTML output matching the MegaBrain visual identity — dark backgrounds, brand orange, clean typography.

View on GitHub
Design

Use cases

  • HTML presentations and reports
  • Dashboards and data visualizations
  • Any visual output from Cline
When creating any HTML, CSS, or visual output, follow the MegaBrain Design System:

COLORS:
- Page background: #09090b
- Card/section background: #18181b
- Border color: #27272a
- Primary text: #fafafa
- Secondary/muted text: #a1a1aa
- Brand accent (headings, highlights): #e4530a
- Success/code: #22c55e
- Link hover: #e4530a

TYPOGRAPHY:
- Font stack: system-ui, -apple-system, 'Segoe UI', sans-serif
- Heading sizes: h1=2.5rem, h2=1.75rem, h3=1.25rem
- Heading weight: 700-800, letter-spacing: -0.02em
- Body line-height: 1.65
- Monospace: 'JetBrains Mono', 'Fira Code', monospace

COMPONENT PATTERNS:
- Cards: background #18181b, border 1px solid #27272a, border-radius 16px, padding 24px
- Badges: border-radius 999px, padding 4px 12px, border 1px solid #3f3f46, font-size 0.75rem
- Code blocks: background #09090b, border-radius 12px, padding 16px, font-family monospace
- Buttons (primary): background #e4530a, color white, border-radius 8px, padding 10px 20px
- Dividers: border-color #27272a, opacity 0.6

LAYOUT:
- Max content width: 1200px, centered
- Section padding: 48px 24px
- Grid gap: 16-24px
- Always use CSS Grid or Flexbox, never tables for layout

CHARTS (when using Chart.js or similar):
- Background: transparent
- Grid lines: rgba(255,255,255,0.06)
- Tick color: #71717a
- Dataset colors in order: #e4530a, #3b82f6, #22c55e, #a855f7, #f59e0b
- Legend label color: #a1a1aa

Always produce self-contained HTML with all CSS inline or in a <style> tag. No external CSS frameworks unless Chart.js CDN for charts.

McKinsey 7-Step Problem Solving + Pyramid Principle

Structures data analysis using McKinsey's 7-step problem solving framework and presents findings using the Pyramid Principle (SCQA + grouped arguments).

View on GitHub
Analysis

Use cases

  • Business analysis and data reports
  • Strategy decks
  • Executive summaries
When analyzing data or creating presentations, apply the McKinsey 7-Step Problem Solving framework combined with the Pyramid Principle.

ANALYSIS FRAMEWORK McKinsey 7 Steps:

Step 1 DEFINE THE PROBLEM
- Restate the problem in one crisp sentence
- Identify: Who is the decision-maker? What decision do they need to make? By when?
- Define what "solved" looks like (success metrics)

Step 2 STRUCTURE THE PROBLEM (Issue Tree)
- Break the problem into MECE sub-questions (Mutually Exclusive, Collectively Exhaustive)
- Use a logic tree: main question 3-5 branches sub-branches
- Label each branch as a hypothesis to test

Step 3 PRIORITIZE ISSUES
- Rank sub-questions by: (a) impact on the answer, (b) ease of analysis
- Focus 80% of effort on the 20% of issues that drive the answer
- State explicitly what you are NOT analyzing and why

Step 4 BUILD THE WORK PLAN
- For each priority issue: what data is needed, what analysis, what output
- Assign a "ghost slide" concept to each: what would the chart/table look like if the hypothesis is true?

Step 5 CONDUCT ANALYSES
- Run the analyses on the actual data provided
- For each finding: state the so-what, not just the what
- Flag data quality issues or gaps explicitly

Step 6 SYNTHESIZE FINDINGS
- Group findings into 3-5 key messages (no more)
- Each message must be: action-oriented, specific, and supported by at least one data point
- Identify the "single most important insight" the one thing the decision-maker must know

Step 7 DEVELOP RECOMMENDATIONS
- Lead with the recommendation, not the analysis
- Structure as: Recommendation Evidence Risk/Mitigation Next step
- Be concrete: who does what, by when, measured how?

---

PRESENTATION STRUCTURE Pyramid Principle (Barbara Minto):

Opening SCQA Framework:
- Situation: What is the stable context everyone agrees on?
- Complication: What has changed or what is the problem?
- Question: What question does this raise for the audience?
- Answer: Your top-line recommendation (state it upfront, always)

Body Grouped Arguments:
- Present 3 main supporting arguments (never 2, rarely more than 5)
- Each argument must directly support the top-line answer
- Support each argument with data, not assertions
- Use the "So what?" test on every slide/section

Slide/Section Titles:
- Write action titles, not topic titles
- Bad: "Revenue Trends" Good: "Revenue grew 23% driven by Q4 enterprise wins"
- Every title should be a complete sentence stating the insight

Chart Principles:
- One message per chart
- Title states the insight the chart proves
- Highlight the key data point (bold, color, annotation)
- Remove all chart junk (unnecessary gridlines, legends when obvious, 3D effects)

---

OUTPUT FORMAT:
When creating an HTML presentation, structure it as:
1. Title slide: problem statement + recommendation in one sentence
2. Executive Summary: SCQA in 4 bullet points
3. Key Finding slides (one per main argument, 3-5 total)
4. Recommendation slide: action + owner + timeline + success metric
5. Appendix: supporting data, methodology, assumptions