Knowledge Maps

Knowledge Maps accumulate insights from multiple exploratory sessions, helping BugBrain improve test quality and coverage over time.

What Is a Knowledge Map?

A Knowledge Map is a persistent record of everything learned about a target application:

• Navigation Graph — Pages discovered, links between them, user flows
• Coverage Tracking — Which pages/features have been tested, which haven’t
• Risk Heatmap — Areas with frequent bugs, potential problem zones
• Feature Inventory — Forms, buttons, flows, and interactions discovered
• Pattern Library — Reusable UI patterns and their expected behavior

One Knowledge Map per target URL — All exploratory sessions for the same domain add to the map.

Building Knowledge Over Multiple Sessions

Session 1: 50% Confidence

Run first exploratory session on example.com

• Discovers 20 pages
• Tests basic flows
• Finds 5 bugs
• Knowledge confidence: 50%

Session 2: 75% Confidence

Run second session on same URL

• Discovers 3 new pages (17 previously known)
• Tests more interaction combinations
• Finds 8 bugs (3 new, 5 similar to session 1)
• Knowledge confidence: 75%

Session 3: 95% Confidence

Run third session

• Discovers 1 new page (22 previously known)
• All major flows covered
• Finds 12 bugs (6 new, 6 repeat areas)
• Knowledge confidence: 95%

Result: Each subsequent session finds more relevant bugs because the AI understands the application structure better.

Knowledge Map Dashboard

Navigation Graph

Visual representation of page structure and user flows:

[Homepage]
    ↓
[Product Listing] ← [Search Results]
    ↓
[Product Detail] → [Reviews]
    ↓
[Add to Cart]
    ↓
[Checkout] → [Payment]
    ↓
[Order Confirmation] → [Order History]

Shows:

• All discovered pages
• Navigation paths between pages
• Flow entry/exit points
• Unreachable pages (dead links)

Use for:

• Understanding application structure
• Identifying missing flows
• Planning test case creation
• Documenting user journeys

Coverage Metrics

Track which parts of your application have been tested:

Coverage Formula:

Coverage = (Pages Tested / Pages Discovered) × 100

Risk Heatmap

Areas where bugs are frequently found:

Dashboard (8 bugs)       ████████ HIGH RISK
Payment Flow (6 bugs)    ██████ MEDIUM-HIGH
Profile Settings (4)     ████ MEDIUM
Search (2)               ██ LOW
Homepage (0)             LOW RISK

Use for:

• Prioritizing remediation (fix high-risk areas first)
• Allocating test resources
• Planning exploratory session focus areas

Using Knowledge to Generate Better Test Cases

Once a Knowledge Map has 75%+ confidence, it can improve AI-generated test cases:

1. Auto-Generate Test Cases

   • AI reads the Knowledge Map
   • Generates test cases for discovered flows
   • Includes edge cases the agent found bugs in
   • 3-5x more effective than random generation

2. Smarter Persona Selection

   • Maps shows which flows require authentication
   • Suggests creating personas for role-based flows
   • Identifies MFA/OTP requirements automatically

3. Better Bug Reproduction Steps

   • Maps includes exact navigation paths
   • AI-generated test cases use real paths
   • Higher success rate when reproducing bugs

Knowledge Map Features

Pages Inventory

Example:

Homepage (/index)
  - 8 links found
  - 2 forms found
  - Login button (header)
  - Search bar (header)

Product Listing (/products)
  - Filters: Category, Price, Rating
  - Pagination: 1-20 of 500+ products
  - Sort options: Relevance, Price, Rating

Product Detail (/products/[id])
  - Related products section
  - Customer reviews
  - Add to cart button
  - Share buttons

Interaction Patterns

Buttons & Links:

• Location: Navigation, header, product card, footer
• Actions: Submit, Login, Add to Cart, Save, Delete
• Expected behavior: Page navigation, modal open, form submission

Forms:

• Location: Login, signup, search, filters, checkout
• Fields: Text, email, password, dropdown, checkbox, radio
• Validation: Required fields, email format, password strength

Dynamic Content:

• AJAX search results
• Lazy-loaded pagination
• Infinite scroll
• Modal dialogs
• Tab switching

Risk Areas Detected

High-Risk Patterns:

• Manual user input (forms, search) — injection vulnerabilities
• Financial transactions (payment) — money loss risk
• User data (profile) — privacy/security risk
• Authentication (login) — access control issues

Domain-Specific Knowledge Maps

Knowledge Maps understand application type, enabling targeted testing:

The AI agent adapts its testing strategy based on detected application type.

Maintaining Accurate Knowledge Maps

When to Rebuild Knowledge Maps

Keep current map:

• Minor bug fixes
• UI improvements (same structure)
• Content updates

Rebuild map:

• Major redesign (new page structure)
• New feature launched
• Reorganized navigation
• Application rebranding

To rebuild:

1. Go to Settings → Knowledge Maps
2. Select target URL
3. Click Reset Map (archive old map)
4. Run new exploratory session to build fresh knowledge

Best Practices

1. Run sessions regularly — Weekly/monthly to keep map fresh
2. Document changes — Note when you rebuild maps (UI changes, releases)
3. Review coverage — Check if covered pages match your release notes
4. Act on risk heatmaps — Investigate high-risk areas after bugs found

Exporting Knowledge Maps

Generate documentation for team reference:

PDF Report:

• Navigation graph (visual diagram)
• Pages inventory with descriptions
• Coverage metrics
• Risk heatmap
• Interaction patterns found
• Recommended test cases

Markdown/HTML:

• Navigation map as text/HTML
• Feature inventory tables
• Risk assessment
• Suitable for wiki/documentation sites

JSON:

• Raw Knowledge Map data
• Pages, links, forms, interactions
• Coverage and risk metrics
• For tooling integration

Use exported Knowledge Maps to:

• Share with new team members
• Update project documentation
• Plan feature rollouts
• Justify testing resource allocation