exploratory-testing-advanced

This skill provides a structured framework for advanced exploratory testing, shifting the practice away from random clicking toward disciplined, cognitive investigation. It centers on Session-Based Test Management (SBTM), which organizes exploration into time-boxed charters with defined missions, scope, and specific quality focus areas. By utilizing established heuristics such as SFDIPOT for coverage analysis and FEW HICCUPPS for identifying problems, agents can systematically evaluate software reliability, data integrity, and operational usability. The skill also incorporates various test tours—such as the Money tour for revenue-impacting features or the Bad Neighborhood tour for high-risk areas—to ensure comprehensive coverage of complex systems.

• Employs SBTM methodology to manage exploration sessions (Charter, Explore, Note, Debrief) with 45-90 minute time-boxes.
• Leverages SFDIPOT (Structure, Function, Data, Interfaces, Platform, Operations, Time) to assess quality criteria and identify coverage gaps.
• Uses FEW HICCUPPS (Familiar, Explainable, World, History, Image, Comparable, Claims, Users, Product, Purpose, Statements) as consistency oracles to recognize and categorize bugs.
• Implements 12 distinct test tour strategies to navigate applications through different lenses, including Business District, FedEx, and Garbage Collector tours.
• Provides a robust note-taking template for real-time documentation of observations, bugs, questions, and session coverage metrics.
• Supports agent-assisted exploration, enabling collaboration between human navigators and AI drivers for tasks like edge case generation, visual testing, and pattern recognition.

Usage notes and practical constraints:

• Ideal for investigating new features, finding bugs that automated suites might miss, learning unfamiliar systems, and risk discovery during early development phases.
• Requires active engagement; exploration is framed as skilled, structured thinking rather than unguided execution.
• Facilitates session sharing via agent memory, allowing teams to track bug clusters, re-use charter templates, and build a library of heuristic results.
• Best utilized in pairing patterns such as Driver-Navigator or Strong-Style, where the agent handles repetitive variations while the human focus remains on high-level strategy and intuition.
• Designed for integration within the broader Agentic Quality Engineering (AQE) fleet, coordinating specialized agents like qe-flaky-test-hunter and qe-visual-tester.