Guide · Agent skill
How to use the Test Coverage Gap Analyzer agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The coverage skill is an instruction file for an AI agent. When you ask what your app is not testing, it tells the agent to map your pages, components and API endpoints, compare them against your Playwright spec files, and report the gaps in priority order. This guide is for developers and QA engineers who maintain a web app with Playwright end-to-end tests and want a clear list of what to test next.
Test Coverage Gap Analyzeralirezarezvani/claude-skills · MIT Download the skill Quick start
- 1Download the ZIP. Download the skill ZIP file from this page.
- 2Unzip the skill. Unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ in your project for team use.
- 3Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.
- 4Make sure your suite is ready. Confirm the project has *.spec.ts or *.spec.js Playwright files that the agent can read.
- 5Ask in plain language. Ask something like "What's not tested in this app?" so the agent maps routes, components and endpoints.
- 6Review the matrix and plan. Read the coverage matrix and the prioritized gap list, and check the suggested test counts.
- 7Decide on generation. Answer Yes, No or Pick specific when asked, and review any tests generated through /pw:generate.
What the skill makes the agent do
The skill runs in six steps. First the agent maps the application surface: routes and pages, interactive components, API endpoints and critical user flows such as auth, checkout and onboarding. Then it scans all *.spec.ts and *.spec.js files to see which pages are opened with page.goto(), which components are reached through locators, and which API endpoints are mocked or hit.
The agent then builds a coverage matrix showing each area, route, test count and status, ranks the uncovered areas from critical to low, and suggests how many tests each gap needs. The output includes a coverage percentage estimate.
Install the skill
Download the ZIP file from this page. Unzip it into one of two places: ~/.claude/skills/ to make the skill available to you across projects, or the .claude/skills/ folder inside a single project to share it with that project.
You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- Personal: unzip into ~/.claude/skills/
- Single project: unzip into .claude/skills/ in the project root
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
Trigger the skill with natural requests
The skill's description lists phrases that activate it. You do not need to name the skill. Ask in plain language using wording such as the phrases below.
- "What's not tested in this app?"
- "Show me the test coverage gaps."
- "Which areas have missing tests before the release?"
- "Give me a coverage report for the Playwright suite."
- "What needs testing first?"
Read the output and decide what to do next
The coverage matrix uses a status column with three values: covered, partial and missing. A partial status comes with a note on what is absent, such as error states. The priority list runs from Critical (auth, payment, core features) through High (user-facing CRUD, search, navigation) and Medium (settings, preferences, edge cases) to Low (static pages such as about and terms).
The recommended test plan lists each gap with a test count, a suggested template from a templates/ folder and an effort estimate of quick, medium or complex. Use this list to decide what to write first.
Optionally generate the missing tests
After the report, the agent asks whether to generate tests for the top N gaps, with the choices Yes, No or Pick specific. If you say yes, it calls the /pw:generate command for each gap using the recommended template. That command comes from the same playwright-pro package and is not included in the file on this page.
Because the agent asks first, you stay in control of which gaps get generated. Review the generated spec files as you would any code change.
What you need before you start
The skill is built around Playwright. Your project needs a Playwright test suite with *.spec.ts or *.spec.js files, and the agent must be able to read your project files. The optional generation step needs the /pw:generate command. The skill also refers to a templates/ folder that is not included in the file provided here, so the template suggestions depend on that folder existing in your setup.
- A Playwright suite with *.spec.ts or *.spec.js files
- An agent that can read the project files
- Optional: the /pw:generate command from the playwright-pro package
- Templates from a templates/ folder, which are not included in the provided SKILL.md
Safety notes
The skill runs no scripts and uses no network access, according to its listing. It works by reading your route definitions and spec files. The only action that changes your project is the optional test generation, and the skill asks for confirmation before running it.
Tips for useful results
The coverage percentage is an estimate built from routes and page.goto() calls. It is not line or branch coverage from instrumented code, so do not report it as a measured coverage number. Use the matrix to find gaps, not to set a precise target.
Name your critical flows in the request, such as checkout or login, so the agent weights them correctly. Teams inheriting an app with unclear tests will get the most from the first run.
- Treat the percentage as a rough guide, not a measurement
- Mention business-critical flows so they rank as Critical
- Check that your spec files use page.goto() and locators, since the scan relies on them
What people use it for
- Finding untested auth pages such as register and forgot-password before a release
- Listing which dashboard error states have no tests in an inherited app
- Planning which areas a QA team tests first in a sprint
- Getting a prioritized list of gaps to hand to a developer as a backlog
Frequently asked questions
How do I install the coverage skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into .claude/skills/ inside a project. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
Does the coverage skill need Playwright?
Yes. It scans *.spec.ts and *.spec.js files and uses page.goto() calls and locators, so it is built around Playwright. It has no support for other test frameworks.
Is the coverage skill free?
The skill is published under the MIT licence. This page does not list pricing, so check the licence file in the repository for the terms.
Does the skill measure line or branch coverage?
No. The percentage is estimated from routes and page.goto() calls, not from instrumented code. Treat it as a rough guide to gaps, not a measured coverage number.
Will the skill change my test files automatically?
No. Optional test generation calls the separate /pw:generate command, and the skill asks you before generating tests for the gaps you choose.
What are the alternatives if my project doesn't use Playwright?
This skill only works with Playwright spec files, so it will not scan projects that use other test frameworks. For those projects, look for a tool that matches your framework, since this skill has no support for other frameworks.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you maintain a web app with Playwright tests and want a prioritized list of what is missing, especially when you inherit an app with unclear coverage. Pick something else if your project does not use Playwright, if you need real line or branch coverage numbers, or if you want tests written automatically without review.
Download the skill