Skip to main content

Frontier Exploration

AI-guided discovery of undiscovered application states and paths.

What Is Frontier Exploration?

Frontier exploration uses AI to systematically discover parts of your application that haven't been tested. The AI identifies unexplored states, navigates to them, and maps new user flows automatically.

Use cases:

  • Find hidden or undocumented pages
  • Discover edge cases in user flows
  • Identify dead ends and broken links
  • Build comprehensive test coverage

Explore Frontier

Start automatic discovery from your application's current state.

  1. Open the Knowledge Graph view
  2. Click Explore Frontier
  3. Watch as the AI:
    • Identifies unexplored areas (shown as frontier nodes)
    • Navigates to each undiscovered state
    • Records new paths and interactions
    • Updates the graph in real-time

The AI prioritizes exploration based on:

  • Distance from known states
  • Likelihood of discovering new functionality
  • Importance based on page structure

Explore from Specific State

Target exploration to specific areas of your application.

  1. Click any node in the graph
  2. Select Explore from here in the context menu
  3. The AI deep-dives into that area, discovering:
    • Child states accessible from that node
    • Alternative paths to the same destination
    • Hidden interactions and edge cases

When to use:

  • After adding a new feature
  • To thoroughly test a specific user flow
  • When coverage metrics show gaps in a particular area

Maximizing Coverage

Use exploration data to ensure comprehensive testing.

Review Coverage Metrics

The graph displays coverage indicators:

  • Green nodes: Fully explored states
  • Yellow nodes: Partially explored (some paths untested)
  • Gray nodes: Discovered but unexplored
  • Frontier edges: Paths that lead to unknown states

Identify Gaps

Look for patterns indicating incomplete coverage:

  • Clusters of gray or yellow nodes
  • Long chains without branching
  • Pages with fewer edges than expected

Target Specific Flows

Focus exploration on critical paths:

  1. Select a starting node (e.g., "Login")
  2. Choose Explore from here
  3. Review discovered paths to key destinations (e.g., "Checkout")
  4. Generate tests for the discovered flows

Using Exploration Data

Convert discoveries into actionable testing artifacts.

Generate Tests

After exploration completes:

  1. Select discovered paths in the graph
  2. Click Generate Tests
  3. AI creates test cases for each flow
  4. Review and add to your test suite

Document Paths

Export exploration results for documentation:

  • Critical user journeys
  • Authentication flows
  • Error recovery paths
  • Edge case scenarios

Identify Problems

Exploration reveals application issues:

  • Dead ends: Pages with no outbound navigation
  • Broken links: Paths that fail to reach destinations
  • Orphan pages: States unreachable from main navigation
  • Infinite loops: Circular navigation patterns

Exploration Settings

Configure exploration behavior in Settings:

SettingDescriptionDefault
Max depthHow far to explore from starting point5 levels
TimeoutMaximum time per exploration session10 minutes
Skip patternsURL patterns to ignoreAdmin pages
Auth handlingHow to handle login statesPreserve session

Best Practices

Start broad, then focus. Run full frontier exploration first, then target specific areas with gaps.

Explore after changes. Run exploration when you add new features or modify navigation.

Review before generating tests. Check discovered paths for relevance before creating test cases.

Track coverage over time. Use the coverage metrics to ensure exploration keeps pace with development.