expand-coverage

name	expand-coverage
description	Analyze test coverage gaps on changed or specified files, spawn tester agents to write tests, and verify improvement. Use when you need to get coverage above 80% on new code or expand coverage on existing modules.
allowed-tools	Bash, Read, Edit, Write, Glob, Grep, Task
skills	["dancing-bear-rules"]

Expand Coverage

Find coverage gaps, write tests, and verify improvement.

When to Use

• After writing new code that needs tests
• User says "add tests" or "improve coverage"
• Coverage below 80% on new/changed files
• As part of /review-fix Phase 2

Step 1: Identify Gaps

# Full coverage report (all domains)
make cov

# Coverage for a specific domain
PYTHONPATH=. python3 -m coverage run -m unittest discover -s tests/<domain> -t . -q
python3 -m coverage report -m --include="<domain>/*"

# Lint-level complexity/smell signals (not coverage, but flags files needing tests)
~/.qlty/bin/qlty check <domain>/

Step 2: Prioritize

Order by impact:

1. New files with 0% coverage — need test skeletons
2. Changed files below 80% — need gap tests
3. Critical paths below 90% — auth, credentials, API clients

Step 3: Write Tests

For each file needing coverage:

1. Read the source to understand what to test
2. Read domain conftest.py for available factories and fixtures:
   • tests/conftest.py — global fixtures, AWS popup prevention
   • tests/<domain>/conftest.py — domain-specific factories
   • tests/helpers/ — mock libraries (github_mocks.py, atlassian_adf_mocks.py)
3. Write tests following project conventions

Test Conventions (Mandatory)

• Use make_* factories from conftest, never construct dicts manually
• Use conftest constants (TEST_DATE_CREATED, DEFAULT_STATUS, etc.)
• Patch where the name is used, not where it's defined
• Specific assertions: assertEqual, assertIn, assertIsInstance
• @dataclass for test fixtures

Coverage Strategy

1. Happy path first
2. Edge cases (None, empty, invalid inputs)
3. Error paths (failures, timeouts, exceptions)
4. Mock all external services

For large coverage expansion, spawn tester agents:

Task(subagent_type="tester", prompt="Write tests for <domain>/<module>.py targeting 80%+ coverage. Read tests/<domain>/conftest.py first for available factories.")

Step 4: Verify

# Run new tests
python3 -m unittest tests.<domain>.test_new -v

# Check coverage improved
PYTHONPATH=. python3 -m coverage run -m unittest discover -s tests/<domain> -t . -q
python3 -m coverage report -m --include="<domain>/*"

# Full suite (no regressions)
make test

Step 5: Report

Output a summary:

## Coverage Expansion Results

| File | Before | After | Tests Added |
|------|--------|-------|-------------|
| domain/module.py | 45% | 82% | 8 |
| domain/cli.py | 0% | 85% | 12 |

Total: N new tests, M files improved

Coverage Targets

Category	Target
New code	80%+
Critical paths (auth, credentials, API)	90%+
Existing code	Improve toward 85%