test-quality-analysis

Analyze test code quality to detect coverage-only tests, test smells, and low-value assertions. Use when asked to "analyze test quality", "find coverage-only tests", "audit our tests", "are these tests valuable", "find test smells", or "which tests should we delete". Scores tests 1-5 on real value and produces prioritized improvement reports.

Jump to install

Repository: fabioc-aloha/ChessCoach
Last source activity: May 5, 2026 at 20:58
Detected SKILL.md language: English
Stars: 1
Forks: 0

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

Back to search Back to installation

type	skill
lifecycle	stable
inheritance	inheritable
name	test-quality-analysis
description	Analyze test code quality to detect coverage-only tests, test smells, and low-value assertions. Use when asked to "analyze test quality", "find coverage-only tests", "audit our tests", "are these tests valuable", "find test smells", or "which tests should we delete". Scores tests 1-5 on real value and produces prioritized improvement reports.
applyTo	*/.test.,/.spec.,/test/,/__tests__/*
tier	standard
currency	2026-04-22T00:00:00.000Z

Test Quality Analysis

Detect tests that exist solely for coverage metrics, identify test smells, and score test value.

When to Use

Auditing test suite quality before a release
Suspecting tests inflate coverage without catching bugs
User asks "are these tests any good" or "find useless tests"

What This Skill Is NOT

Not refactor: That improves production code. This evaluates test code.

Detection Heuristics

Detect the project's test framework by scanning test files for import/require statements and config files, then apply these patterns:

Critical (Score 1-2)

Pattern	Why It's Bad
No assertions — test calls methods but never verifies outcomes	Exercises code without proving it works
Trivial assertions — asserts on constants or always-true conditions	Always passes regardless of behavior
Exception swallowing — catches all errors silently	Hides failures; test can never fail
Self-referential — asserts input equals output when transformation is identity	Tests nothing meaningful

Warning (Score 2-3)

Pattern	Why It's Concerning
Over-mocking — every dependency mocked; no real code executes	Tests the mock setup, not behavior
Coverage touching — calls methods systematically without verifying behavior	Covers lines without testing logic
Weak verification — checks type/shape only, not content	Misses value-level bugs
Missing negative path — only happy path tested	Won't catch error handling bugs

Minor (Score 3-4)

Pattern	Check For
Missing edge cases, incomplete verification, poor naming, test duplication, brittle setup

Value Scoring

Score	Rating	Action
1	Delete	Zero value — pure coverage inflation
2	Rewrite	Valid concept, useless assertions
3	Improve	Some value but assertions too weak
4	Acceptable	Reasonable, minor improvements optional
5	High Value	Catches real bugs, keep as-is

# Test Quality Report
Directory: [path] | Tests: [count] | Date: [date]

## Summary — score distribution table with counts and percentages
## Top Offenders — ranked list of lowest-scoring tests
## Findings by Severity — per-test: file, score, issues, recommendation
## Recommendations — immediate deletes, rewrites, process improvements

# Test Quality Report — tests/services/ | 34 tests

Score distribution: Delete(2) | Rewrite(5) | Improve(12) | Acceptable(11) | High Value(4)

Top Offenders:
  🔴 order.test.ts:12 "should process order" — Score 1
     Zero assertions. → Delete or add assertions for state, DB, events.
  🔴 user.test.ts:45 "should create user" — Score 1
     Only asserts `!= null`. → Rewrite to assert user fields and DB record.
  🟡 auth.test.ts:78 "should validate token" — Score 2
     All deps mocked. → Use real TokenValidator; mock only external APIs.

Actions: Delete 2 zero-assertion tests, rewrite 5 score-2 tests, add assertion check to PR reviews.

# Test Quality Report — tests/services/ | 28 tests

Score Distribution:
  Delete (1): 2 | Rewrite (2): 3 | Improve (3): 4 | Acceptable (4): 12 | High Value (5): 7

Top Offenders:
  🔴 payment.test.ts:23 — Score 1 — Zero assertions → Delete or add state checks
  🔴 user.test.ts:51    — Score 1 — Trivial assertion → Rewrite with field validation
  🟡 auth.test.ts:34    — Score 2 — Over-mocked → Use real TokenValidator
  🟡 order.test.ts:67   — Score 3 — Weak assertion → Assert exact values

Recommendations:
  Immediate: Delete 2 zero-value tests (they provide false confidence)
  Short-term: Rewrite 3 score-2 tests with real assertions
  Process: Add assertion-quality check to PR review checklist

Scenario	Action
Test framework not detected	Ask user for the test framework; apply heuristics generically
Test files use unfamiliar assertion library	Analyze assertion patterns by structure (function calls after act); note reduced confidence
No test files found in specified path	Report "no tests found" and suggest directories to check
Cannot determine production code under test	Analyze the test in isolation; note that bug-detection assessment is limited

❌ Don't	✅ Do Instead
Treat all tests as equally valuable	Score each on real value
Only check assertion count	Check assertion quality and relevance
Delete tests without understanding intent	The idea may be valid even if implementation is bad
Focus only on coverage percentage	High coverage ≠ high quality