Quality engineering for
AI coding agents,
one command away.

Reviewed QA skills for any coding agent, plus agents and role bundles for Claude Code. 329 skills, 73 agents, free and MIT.

Browse 67 plugins
/plugin install qa-starter@testland-qa
Testland for teamsIn development

Test reports, manual runs and regressions in one workspace. Privacy policy

Skills install in the coding agents your team already uses. Agents and role bundles run in Claude Code.

Claude CodeCursorGitHub CopilotGemini CLIWindsurf

From install to answer in three steps.

No setup files and no prompts to learn. Add the marketplace once, install what fits your role, then ask the way you'd ask a teammate.

1

Add the marketplace

One line in Claude Code. Other agents can use skills.sh.

/plugin marketplace add testland/qa

✓ 67 plugins · 11 role bundles

2

Install for your role

qa-starter has what every tester needs, on any stack.

/plugin install qa-starter@testland-qa

✓ 10 plugins · 64 skills · 22 agents

3

Ask in plain words

The agent picks the skill whose trigger matches your request.

write manual test cases for the password reset flow

✓ manual-test-script-author

Install for your role.

Bundles follow how QA teams are staffed. Each request below is picked up by a real skill in the marketplace.

Manual tester7 plugins · 48 skills

“Write manual test cases for the password reset flow”

manual-test-script-author
/plugin install qa-role-manual-tester@testland-qa
Automation engineer9 plugins · 63 skills

“Why does this test fail only on CI?”

flake-axis-bisection
/plugin install qa-role-automation-engineer@testland-qa
Frontend developer8 plugins · 38 skills

“Check this page for accessibility issues”

axe-a11y
/plugin install qa-role-frontend@testland-qa
Backend developer17 plugins · 78 skills

“Did this API change break any clients?”

openapi-contract-diff
/plugin install qa-role-backend@testland-qa
QA lead8 plugins · 40 skills

“Which parts of this release are riskiest?”

risk-matrix
/plugin install qa-role-leadership@testland-qa

Skills for the tools you already run.

129 tools across 11 areas. Each one opens the skill that knows it, with install commands and sources.

Playwright testing skillCypress testing skillSelenium testing skillWebdriverIO testing skillStorybook testing skillChromatic testing skillPercy testing skillaxe-core testing skillWorkbox testing skillAppium testing skillEspresso testing skillXCUITest testing skillMaestro testing skillDetox testing skillFlutter testing skillElectron testing skillQt testing skillFlaUI testing skillWinAppDriver testing skillJest testing skillVitest testing skillpytest testing skillJUnit 5 testing skill.NET testing skillGo testing skillRust testing skillHypothesis testing skillfast-check testing skilljqwik testing skillproptest testing skillStryker testing skillPIT testing skillmutmut testing skillMull testing skillCucumber testing skillBehave testing skillReqnroll testing skillBats testing skillPester testing skillPostman testing skillREST Assured testing skillKarate testing skillPact testing skillSchemathesis testing skillRESTler testing skillOpenAPI testing skillGraphQL testing skillApollo testing skillHasura testing skillProtobuf / Buf testing skillgRPC testing skillWireMock testing skillMSW testing skillFaker testing skillWebSocket testing skillMQTT testing skillLighthouse testing skillk6 testing skillJMeter testing skillghz testing skillChaos Mesh testing skillToxiproxy testing skillOWASP ZAP testing skillTrivy testing skillgitleaks testing skillSemgrep testing skillSnyk testing skillSonarQube testing skillCodeQL testing skillDependabot testing skillRenovate testing skillNuclei testing skillOSV-Scanner testing skillSyft testing skillTruffleHog testing skillCheckov testing skillHelm testing skillAuth0 testing skillKeycloak testing skillOkta testing skillPresidio testing skillTestcontainers testing skillDocker Compose testing skilldbt testing skillGreat Expectations testing skillSoda testing skillFlyway testing skillLiquibase testing skillAtlas testing skillKafka testing skillRabbitMQ testing skillAmazon SQS testing skillCelery testing skillSidekiq testing skillBullMQ testing skillRedis testing skillVarnish testing skillElasticsearch testing skillOpenTelemetry testing skillJaeger testing skillAWS SAM testing skillAzure Functions testing skillCloudflare Workers testing skillLaunchDarkly testing skillOpenFeature testing skillStripe testing skillMailpit testing skillpromptfoo testing skillDeepEval testing skillRagas testing skillGiskard testing skillLangfuse testing skillDeepchecks testing skillEvidently testing skillFairlearn testing skillAllure testing skillCurrents testing skillJaCoCo testing skillcoverage.py testing skillGitHub Actions testing skillGitLab CI testing skillJenkins testing skillKnip testing skillGoogleTest testing skillQEMU testing skillCeedling testing skillUnity testing skillUnreal Engine testing skillGodot testing skill
24 of 129 tools

A QA team that lives in your editor.

Testland packages how experienced testers work into small parts your agent loads only when a request needs them. Every part is reviewed on six dimensions and cites its sources.

329

Skills

One tool or technique each. A skill loads when your request matches its trigger.

73

Agents

One focused job each. They read your code and return findings, not essays.

11

Role bundles

A whole discipline in one install, from manual tester to SDET and QA lead.

Review tests before they merge.

Critic agents read the diff the way a senior SDET would: brittle selectors, fixed sleeps, missing assertions, shared state.

tests/checkout/coupon.spec.tsExample2 findings
test('applies the coupon before tax', async ({ page }) => {
await page.goto('/cart');
await page.getByLabel('Coupon code').fill('SAVE10');
await page.locator('div.actions > button:nth-child(2)').click();

Position-based selector. It breaks as soon as a button is added to the row. Use getByRole('button', { name: 'Apply coupon' }).

await page.waitForTimeout(2000);

Fixed sleep. It is slow on fast runs and flaky on slow ones. Wait for the total with a web-first assertion instead.

await expect(page.getByTestId('order-total')).toHaveText('$89.10');
});

Review summary

Files reviewed4
Selectors checked23
Assertions per test1.8
1 blocking finding. Fix before merge.

Find out why a test is flaky.

e2e-flake-bisector reruns the test while changing one thing at a time (order, timing, environment, shared state) and names the cause.

From the journal: How to Fix Flaky Tests: A Four-Step Triage
checkout.spec.ts › shows the total with taxExample

Timing. Fails when the tax API answers after 300 ms: the test reads the total before the response lands. Wait for the response, not a fixed sleep.

Catch breaking migrations.

schema-diff-reviewer reads the migration and the code that queries it, and returns one verdict per change.

migrations/0142_orders.sqlExample3 changes
ChangeVerdict
DROP COLUMN orders.legacy_ref
Breaking
ALTER TYPE order_status ADD 'held'
Backfill
CREATE INDEX CONCURRENTLY idx_sku
Safe

2 readers still use legacy_ref: billing-export and the orders_daily model. Ship the drop after they move.

Reviewed like code. Cited like research.

Every skill and agent passes the same six-dimension review before it ships. Citations are the hard floor: no source, no merge.

D1

Spec compliance

Valid manifest, frontmatter and file layout.

D2

Scope

One job, no overlap with another skill.

D3

Description

A trigger the agent can match reliably.

D4

Use-case fit

Solves a task testers actually have.

D5

Body quality

Steps, examples and edge cases that work.

D6 · hard floor

Citations

Claims link to docs, standards or papers.

Testland for teams

In development

Automated test reports, manual test runs and regression tracking in one workspace. A separate product, built on the same skills.

One email when it opens, nothing else. Privacy policy

Questions

Good to know

Anything else, open an issue on testland/qa or email hello@testland.io.

What is a skill?

A set of instructions and references that Claude loads when your request matches its trigger. Each skill covers one tool or technique, such as Playwright screenshot tests or k6 load tests.

How is an agent different from a skill?

A skill teaches Claude a method. An agent is a helper that runs one focused job, like reviewing a database migration, and returns its findings as a table.

How do I install the marketplace?

In Claude Code, run /plugin marketplace add testland/qa once, then install a role bundle or any single plugin, for example /plugin install qa-starter@testland-qa.

Do I need Claude Code?

For agents and role bundles, yes. Skills alone also install into Cursor, Copilot, Gemini, Windsurf and 20+ other agents with npx skills add testland/qa.

How is each component reviewed?

Every pull request is checked against six dimensions, D1 to D6. Concrete claims need a source cited where they are made, and uncited claims are rejected.

Can I install just one plugin?

Yes. Add the marketplace, then install any plugin by name, for example /plugin install qa-accessibility@testland-qa.

Is it free?

Yes. testland-qa is open source under the MIT license.

Start with qa-starter.

10 plugins every tester needs on any stack: test planning, test data, environments, reporting, flaky tests, bug reports and exploratory testing.

/plugin install qa-starter@testland-qa