Hubs Skill

Find hub notes - highly connected notes that serve as central points in your knowledge graph.

Purpose

Hub notes are the opposite of orphans. They have many connections (backlinks + forward links) and serve as:

• Knowledge entry points (MOCs, indexes)
• Central concepts (frequently referenced ideas)
• Active topics (currently being worked on)
• Important entities (key people, projects, technologies)

Finding hubs helps you:

• Identify your vault's "center of gravity"
• Understand what topics are most important to you
• Find natural MOC (Map of Content) candidates
• Discover over-connected notes that might need splitting

When to Use

Invoke when you want to:

• Find most connected notes: "hubs" or "hub notes"
• Discover vault structure: "most connected" or "popular notes"
• Find MOC candidates: "central notes" or "important notes"
• Audit note importance: Part of vault analysis

Process

1. Call MCP Tool

Call: mcp__flywheel__find_hub_notes
Parameters: { min_links: 5 }  # Default threshold

2. Categorize Hubs

Group hubs by type for better analysis:

By Connection Type:

Balanced hubs (many backlinks + forward links):
  - Project X: 45 backlinks, 23 forward links (good MOC)
  - Technology Y: 34 backlinks, 28 forward links (central concept)

Backlink hubs (many backlinks, few forward links):
  - Daily Habit: 300+ backlinks, 2 forward links (referenced entity)
  - Person Name: 67 backlinks, 3 forward links (person)

Forward link hubs (many forward links, few backlinks):
  - Tech Index: 3 backlinks, 45 forward links (potential MOC)
  - Work Dashboard: 2 backlinks, 38 forward links (aggregator)

By Hub Strength:

Super hubs (100+ total links): 3
Strong hubs (50-99 links): 12
Medium hubs (20-49 links): 45
Weak hubs (5-19 links): 96

3. Display Report

Hub Notes Report
═══════════════════════════════════════════════

Found 156 hub notes (11% of vault)
Threshold: 5+ total links

TOP HUBS (20 shown):

1. [[Daily Habit 1]] - 300+ total links
   ← 300 backlinks | → 2 forward links
   Type: Habit | Daily reference
   Hub strength: SUPER (daily tracking)

2. [[Main Project]] - 68 total links
   ← 45 backlinks | → 23 forward links
   Type: Project | Balanced hub
   Hub strength: STRONG ⭐ (excellent MOC!)

3. [[Core Technology]] - 56 total links
   ← 42 backlinks | → 14 forward links
   Type: Technology | Balanced hub
   Hub strength: STRONG (central tech)

═══════════════════════════════════════════════

BREAKDOWN BY TYPE:
  Habits: 3 (super hubs from daily tracking)
  Technologies: 34 (central tech concepts)
  Projects: 18 (active work)
  People: 23 (frequently mentioned)
  Concepts: 45 (important ideas)

MOC CANDIDATES (balanced hubs):
  Main Project: 45 ← | 23 → (⭐ excellent MOC)
  Core Technology: 34 ← | 28 → (⭐ good MOC)

Options:
1. Show full list (all 156 hubs)
2. Filter by type/strength
3. Find potential MOCs (balanced hubs)

Implementation Details

Hub Criteria

A note is a hub if:

• Total links (backlinks + forward links) >= min_links
• Default threshold: 5 links
• User can adjust: "hubs with 20+ links"

Hub Strength Classification

def classify_hub_strength(total_links):
    if total_links >= 100:
        return "SUPER"
    elif total_links >= 50:
        return "STRONG"
    elif total_links >= 20:
        return "MEDIUM"
    else:
        return "WEAK"

Hub Balance Score

Measures how balanced backlinks vs forward links are:

def hub_balance_score(backlinks, forward_links):
    if backlinks == 0 or forward_links == 0:
        return 0  # Completely unbalanced

    ratio = min(backlinks, forward_links) / max(backlinks, forward_links)
    return ratio * 100  # 0-100%

Balance interpretation:

• 80-100%: Excellent MOC candidate
• 50-79%: Good MOC candidate
• 20-49%: Could improve as MOC
• <20%: Unbalanced (entity vs aggregator)

Related Skills

• orphans: Opposite of hubs (zero connections)
• backlinks: Show specific note's connections
• clusters: Find groups of related hubs
• health: Overall vault metrics (hub % is key metric)

Performance

• MCP call: ~300-500ms for large vaults
• Categorization: ~100ms for 156 hubs
• Total: Usually <1 second