Mutation Testing

Run cargo-mutants, analyze missed mutants, write tests for new code, verify fixes.

Workflow

1. Run: Scope to changed files or specified targets
2. Analyze: Focus on new/changed code missed mutants (ignore pre-existing)
3. Fix: Add tests for missed mutants in new code
4. Verify: Re-run on fixed functions to confirm 0 missed

Commands

# Scoped to specific files (recommended for incremental work)
cargo mutants --no-shuffle -j4 -f src/config.rs -f src/volatility.rs

# Scoped to specific functions
cargo mutants --no-shuffle -j4 -F "function_name"

# Full run (slow, 1000+ mutants)
cargo mutants --no-shuffle -j4 -- --lib

# Check results
cat mutants.out/missed.txt | grep "target_file.rs"
cat mutants.out/caught.txt | wc -l

Options

Option	Description
-f <FILE>	Limit to specific source file
-F <PATTERN>	Filter by function name pattern
-E <PATTERN>	Exclude by function name
-j <N>	Parallel jobs (default: auto)
--no-shuffle	Deterministic ordering
-- --lib	Only run library tests

Result Interpretation

Result	Meaning	Action
caught	Test detected mutation	None (good)
missed	Test didn't detect	Add/improve test
unviable	Compile error from mutation	Ignore
timeout	Test too slow	Optimize or skip

Triage Strategy

• Pre-existing missed mutants in main.rs (CLI entry point): low priority, hard to unit test
• Output formatting functions (report.rs, cli_output.rs): low priority unless logic-heavy
• Business logic (config.rs, balance.rs, analyzer.rs): high priority
• New code: always fix missed mutants before release