/coverage-gaps - Coverage Gap Detector

Identify untested code paths introduced by recent changes and generate focused test cases to close gaps. Cross-references diffs against coverage data to find precisely what needs testing.

System Prompt

EXECUTION DIRECTIVE: When the user invokes this command, you MUST:

1. IMMEDIATELY execute - no questions first
2. Load the agent coverage-gap-detector
3. Detect changes via git diff against base
4. Run coverage if available, or analyze test files statically
5. Map gaps between changed code and test coverage
6. Classify gaps by risk level
7. Generate focused test cases for each gap
8. Output gap analysis with generated tests

---

Usage

# Default: diff against main
/coverage-gaps

# Specific base
/coverage-gaps --base develop

# Only critical/high risk gaps
/coverage-gaps --min-risk high

# Generate tests and write to files
/coverage-gaps --output-dir tests/generated/

# Run generated tests to validate
/coverage-gaps --run-generated

# Specific test framework
/coverage-gaps --framework jest

Options

Option	Description	Default
--base	Base reference for diff	main
--min-risk	Minimum risk level to report	medium
--generate-tests	Auto-generate test code	true
--run-generated	Run generated tests to validate	false
--output-dir	Write test files to directory	none
--framework	Test framework (pytest, jest, etc.)	auto-detect

Related Commands

• /test-generate - Generate tests for specific functions
• /test-coverage - View overall coverage metrics
• /flaky-tests - Ensure generated tests are stable
• /bug-scan - Identify which gaps are in bug-prone code

Success Output

COMMAND COMPLETE: /coverage-gaps
Base: main | Head: HEAD
Files Changed: 8 | Functions Modified: 12
Gaps Found: 14 (3 critical, 5 high, 4 medium, 2 low)
Tests Generated: 20
Coverage Impact: +7.6% (projected)

Completion Checklist

• Changed files detected
• Coverage data collected or estimated
• Gaps mapped to functions/branches
• Gaps classified by risk
• Tests generated for each gap
• Coverage impact estimated

Failure Indicators

• No changes detected against base
• Cannot run coverage tool
• Test framework not detected

When NOT to Use

• Measuring overall coverage (use /test-coverage)
• Writing tests from scratch (use /test-generate)
• Fixing failing tests (use /ci-analyze)

Anti-Patterns

Anti-Pattern	Problem	Solution
100% coverage target	Diminishing returns	Focus on critical paths
Test implementation details	Brittle tests	Test behavior
Generate without running	Broken tests	Validate generated tests

Principles

• #3 Complete Execution - Detect, classify, generate, validate
• #9 Based on Facts - Gaps from actual coverage data

Full Standard: CODITECT-STANDARD-AUTOMATION.md