behavioral-mutation-analyzer

Behavioral Mutation Analyzer

Overview

This skill systematically analyzes surviving mutants from mutation testing to understand test suite weaknesses and automatically generate improvements. It identifies why mutants survived, categorizes root causes, and produces actionable test enhancements to increase mutation detection rates.

Analysis Workflow

Step 1: Input Collection and Validation

Gather required inputs and verify completeness:

Required Inputs:

• Repository source code (path or files)
• Test suite (test files and framework)
• Mutation testing results (report file or data)

Mutation Result Formats:

• PIT (Java): XML or HTML reports
• Stryker (JavaScript/TypeScript): JSON reports
• mutmut (Python): result files
• Pitest, Infection (PHP), Cosmic Ray, etc.

Validation checklist:

• Source code accessible
• Test suite runnable
• Mutation results parseable
• Mutation tool and version identified

Step 2: Surviving Mutant Extraction

Parse mutation results to identify all surviving mutants:

Extract for each mutant:

• Mutant ID
• Source file and line number
• Mutation operator (e.g., boundary change, negation)
• Original code
• Mutated code
• Status (survived/killed/timeout/error)

Focus on survived mutants: Filter out killed mutants and focus analysis on survivors that indicate test weaknesses.

Step 3: Root Cause Classification

Analyze each surviving mutant to determine why it survived:

Category 1: Insufficient Coverage

Indicators:

• Mutated line not executed by any test
• Mutated method/function never called
• Conditional branch not taken

Analysis:

• Check code coverage data
• Identify uncovered code paths
• Trace execution from test entry points

Example:

// Original
public int calculate(int x) {
    if (x > 0) {
        return x * 2;  // Line 3: Covered
    }
    return 0;  // Line 5: NOT covered
}

// Mutant: Line 5 changed to "return 1;"
// Survives because no test calls calculate() with x <= 0