How often should coverage-review run?

Every Monday or whenever the test suite feels stale; user_invocable.

What if a gap requires code changes?

Mark as a backlog item; do not modify production code in this cycle.

Where is the output saved?

qa/COVERAGE_GAPS.md with sections on what's covered, new tests, overall coverage, and remaining gaps.

coverage-review

Scanned

npx machina-cli add skill ajaywadhara/agentic-sdlc-plugin/coverage-review --openclaw

Files (1)

SKILL.md

968 B

Run every Monday (or manually when test suite feels stale).

STEP 1 — COVERAGE ANALYSIS

Run: npm run coverage Run: npx playwright test --reporter=json

STEP 2 — FIND GAPS

Identify:

Any function in src/ with 0% test coverage
Any screen in docs/SCREENS.md with no corresponding E2E test
Any error state in the wireframes with no test
Any P1 or P2 feature from the PRD with no E2E test
Any route that has never appeared in a Playwright test

STEP 3 — FILL GAPS

For each gap found: write a test. No implementation changes. Gaps that require new feature code are filed as backlog items, not fixed here.

STEP 4 — REPORT

Save to qa/COVERAGE_GAPS.md:

What was covered this week
How many new tests added
Current overall coverage %
Remaining known gaps

Source

git clone https://github.com/ajaywadhara/agentic-sdlc-plugin/blob/main/skills/coverage-review/SKILL.mdView on GitHub

Overview

coverage-review identifies untested areas across the codebase and guides you to fill coverage gaps with tests. It runs weekly (Mondays) or whenever the test suite feels stale, ensuring ongoing test health. The workflow emphasizes test creation only and documents gaps for follow-up.

How This Skill Works

First, run coverage analysis (npm run coverage) and gather results with Playwright (npx playwright test --reporter=json). Then identify gaps such as 0% coverage in src functions, docs screens with no E2E test, untested error states, high-priority PRD features, and unseen routes. Finally, write tests for each gap (no code changes) and save the summary to qa/COVERAGE_GAPS.md.

When to Use It

Weekly Monday coverage health check
After merging features or PRD high-risk areas
When coverage reports show gaps or new 0% coverage in src
Before a QA cycle to tighten test coverage
After updating docs screens or wireframes to ensure E2E tests exist

Quick Start

Step 1: Run coverage analysis (npm run coverage) and collect Playwright results (npx playwright test --reporter=json)
Step 2: Identify gaps in src, docs/SCREENS.md, wireframes, PRD features, and routes
Step 3: Write tests for each gap and save a weekly report to qa/COVERAGE_GAPS.md

Best Practices

Run npm run coverage and npx playwright test --reporter=json to gather data
Prioritize gaps linked to P1/P2 features from the PRD
Limit changes to test creation; no production code changes during this step
Document weekly results in qa/COVERAGE_GAPS.md
Flag gaps that require new feature code for backlog

Example Use Cases

0% test coverage found in a function under src/
A screen listed in docs/SCREENS.md has no E2E test
An error state in the wireframes has no test
A P1/P2 feature from the PRD lacks an E2E test
A route never appears in any Playwright test

Frequently Asked Questions

Add this skill to your agents