TestMu AI Testing Skills

Equip coding assistants to author and run automated tests on the TestMu AI cloud.

Stars
★ 369
Last updated
1d ago
License
MIT
Primary language
Python

At a glance

Works with
Platform-specificClaude Code
You'll need
Latest stable Node.jsnpmTestMu AI accountLT_USERNAMELT_ACCESS_KEYShell / CLINetwork accessLocal filesystem
Typical use
A QA engineer using Claude Code, Cursor, Copilot, or Gemini CLI wants Playwright login tests that run against Chrome and Firefox in the TestMu AI cloud.
Main limitation
Cloud execution depends on a TestMu AI account plus LT_USERNAME and LT_ACCESS_KEY, creating a material dependency on the provider.

What does this agent do, and when should you use it?

This repository is a collection of TestMu AI (formerly LambdaTest) test-automation skills for coding assistants, not a standalone AI agent. Its skills are installed through `npx agentskillsforall` and are documented for assistants including Claude Code, GitHub Copilot, Cursor, and Gemini CLI. Named packages include Selenium, Playwright, and Cypress skills, while the broader collection is described as covering major automation frameworks across more than 15 languages. A user asks for tests in natural language, the coding assistant authors the relevant test code, and TestMu AI credentials enable execution on the provider's testing cloud. Results are viewed in the TestMu AI dashboard, with a separate TestMu AI Tunnel required when the application under test is hosted locally.

The user installs the repository or an individual selenium-skill, playwright-skill, or cypress-skill with npx agentskillsforall add. The installed skill supplies framework-specific testing guidance to a coding assistant, which can respond to natural-language requests by producing automation code for technologies such as Selenium, Appium, and Playwright; the supplied examples also cover JUnit 5, Mockito, and GitHub Actions CI. For cloud execution, the environment supplies LT_USERNAME and LT_ACCESS_KEY, and tests run through TestMu AI's cloud of more than 10,000 real devices and 3,000 browser environments. Test results are exposed in the TestMu AI dashboard. Locally hosted targets require the OS-specific TestMu AI Tunnel and tunnel: true in the test capabilities.

  1. A QA engineer using Claude Code, Cursor, Copilot, or Gemini CLI wants Playwright login tests that run against Chrome and Firefox in the TestMu AI cloud.
  2. A team maintaining Selenium suites wants its coding assistant to generate test code aligned with the TestMu AI execution environment.
  3. A Java service developer needs JUnit 5 payment-service tests with Mockito and a GitHub Actions CI configuration.
  4. A mobile QA team wants to author Appium automation for execution against TestMu AI's real-device cloud.
  5. A development team testing an application available only on a workstation or private network can connect it through TestMu AI Tunnel and enable tunnel: true.

How do you install or deploy this agent?

Prerequisites are the latest stable Node.js and npm, a supported AI coding assistant, and a TestMu AI account with a username and access key.

Install the repository's skills:

git clone https://github.com/LambdaTest/agent-skills && cd agent-skills
npx agentskillsforall add https://github.com/LambdaTest/agent-skills.git

Alternatively, install one named skill:

npx agentskillsforall add https://github.com/LambdaTest/agent-skills.git --skill selenium-skill
npx agentskillsforall add https://github.com/LambdaTest/agent-skills.git --skill playwright-skill
npx agentskillsforall add https://github.com/LambdaTest/agent-skills.git --skill cypress-skill

Set credentials on macOS or Linux:

export LT_USERNAME="YOUR_USERNAME"
export LT_ACCESS_KEY="YOUR_ACCESS_KEY"

Set credentials on Windows:

set LT_USERNAME="YOUR_USERNAME"
set LT_ACCESS_KEY="YOUR_ACCESS_KEY"

How do you use this agent?

List the available skills with:

npx agentskillsforall list https://github.com/LambdaTest/agent-skills.git

After installing the required skill and setting LT_USERNAME and LT_ACCESS_KEY, give the coding assistant a request such as:

Write Playwright tests for the login page and run them on TestMu AI cloud (Chrome + Firefox)

Or:

Create JUnit 5 tests for the payments service with Mockito and GitHub Actions CI

View completed test results in the TestMu AI dashboard. If the target application is only locally accessible, configure the appropriate TestMu AI Tunnel and add tunnel: true to the capabilities.

What are this agent's strengths and limitations?

Pros
  • Named selenium-skill, playwright-skill, and cypress-skill packages can be installed individually instead of requiring the whole collection.
  • A single agentskillsforall installation flow is documented across Claude Code, GitHub Copilot, Cursor, and Gemini CLI.
  • The skills connect test authoring to a cloud offering more than 10,000 real devices and 3,000 browser environments.
  • Documented tunnel paths for Windows, macOS, and Linux extend testing to applications that are not publicly reachable.
Limitations
  • Cloud execution depends on a TestMu AI account plus LT_USERNAME and LT_ACCESS_KEY, creating a material dependency on the provider.
  • Installation requires the latest stable Node.js, npm, network access, and the separate agentskillsforall CLI.
  • Testing locally hosted applications adds tunnel setup and a required tunnel: true capability change.
  • The supplied material does not provide a complete per-skill framework and language-version compatibility matrix, nor does it state cloud-testing prices or free-tier limits.
  • The Rook skill for testing owned AI agents is not delivered through this repository's agentskillsforall flow; it uses a separate repository and installer and requires Node.js 22 or newer.

How does this agent compare with similar options?

Key facts side by side with the most closely related agents.

Agent Source review Stars Updated Language Full support on
TestMu AI Testing Skills This agent 37 · Major gaps ★ 369 1d ago Python Claude Code
Steel Browser 29 · Major gaps ★ 7.7k 2d ago TypeScript
Supercov 75 · Some gaps ★ 70 4d ago Rust Codex · Claude Code
Carbon Code 64 · Some gaps ★ 196 1mo ago TypeScript OpenAI API

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Major gaps
37/ 100 5-point scale 1.9 / 5
Trust 7/29
Reliability 6/14
Adaptability 9/18
Convention 10/18
Effectiveness 4/13
Verifiability 1/8
Why each dimension lost points
Trust7 / 29 · 1.2/5

The README instructs users to supply a username and access key through environment variables and says tests run on the TestMu AI cloud with results shown in its dashboard, providing limited permission and data-flow visibility. It does not define least-privilege scopes, transmitted data, retention, log redaction, pre-action confirmation, revocation, or recovery. Installation uses unpinned npx and remote Git content without dependency auditing or supply-chain controls. Repository naming, copyright, support contacts, and linked properties provide reasonably clear self-attribution, although publisher identity remains independently unverified.

Reliability6 / 14 · 2.1/5

The README is mostly internally coherent about installation, credential setup, skill selection, cloud execution, and the brand transition, and it identifies Node.js, npm, an AI assistant, and an account as dependencies. Deductions reflect the absence of skill source, dependency manifests, a compatibility matrix, pinned versions, failure-handling procedures, and concrete error messages. A support address and requested bug-report environment details are useful but do not constitute runtime diagnostics.

Adaptability9 / 18 · 2.5/5

The material identifies testers and development teams and names Selenium, Playwright, Cypress, JUnit, several coding assistants, CI, and local-tunnel scenarios. It also provides credential instructions for macOS, Linux, and Windows. However, capability descriptions are largely broad marketing claims, with no per-skill boundaries, supported-version details, exclusions, or conflict rules. Natural-language examples and the --skill option offer only thin trigger guidance and no actual trigger metadata.

Convention10 / 18 · 2.8/5

The README has a clear structure covering prerequisites, installation, execution, local testing, contributions, and support, with understandable commands. Deductions apply because it lacks an FAQ, detailed limitations, an in-repository changelog, compatibility policy, and a stable naming commitment; coexistence of LambdaTest and TestMu AI terminology adds migration ambiguity. The complete MIT license text justifies full license credit. Maintenance and contribution channels are named, but there is no maintainer roster, response commitment, or verified release-responsibility chain.

Effectiveness4 / 13 · 1.5/5

The README describes a usable workflow for generating tests, running them on cloud browsers, and viewing dashboard results, and includes two example requests. No generated artifacts, skill implementation, or comparative results are supplied, so output quality and incremental value over a general coding assistant remain weakly supported. Installation appears short, but account credentials, cloud services, an external installer, and possible tunnel setup impose unquantified costs; pricing, resource consumption, and tradeoffs are not discussed.

Verifiability1 / 8 · 0.6/5

Some operational claims are traceable to concrete README commands and documentation references, while LICENSE corroborates the licensing claim. The supplied evidence contains no skill implementations, tests, dependency manifests, release contents, or independent sources corroborating claims such as “production-grade,” “expert-level,” or support for 15+ languages. Marketing assertions, established facts, and inference are not explicitly separated, so cross-source corroboration and fact/inference separation receive no credit.

Risks and how to mitigate them
  • Not found in source: confirmation before actingTurn on (or add) a confirmation step before it acts, and try it in a sandbox or test environment before real data.
  • Not found in source: dependency securityPin versions and run a dependency audit (npm audit, pip-audit) before installing; prefer running it in a container.
  • Not found in source: rollback or recovery pathBack up first, or work on a git branch or snapshot, so its changes can be undone.
  • Installation relies on an unpinned npx package and remote Git repository; pin a version or commit and inspect the installed content before use.
  • Access keys are configured through environment variables, but credential scope, logging, transmission, retention, and rotation are not documented.
  • Cloud testing may expose code, test data, URLs, or application interactions; avoid real sensitive data until data-flow and privacy terms are established.
  • The evidence contains only a README and license, not skill implementations, dependency manifests, tests, or release contents, so security, output quality, and marketing claims cannot be statically verified.
Evidence confidence: Low Reviewed Sep 11, 2026 Reviewed revision 0de6ebfd44b7
Review evidence README.mdLICENSE
See the full review method →

FAQ

Is a TestMu AI account required?
The documented cloud-execution path requires an account, username, and access key. The material does not establish whether an account is mandatory when a skill is used only to author local test code.
Which coding assistants are supported?
The material explicitly names Claude Code, GitHub Copilot, Cursor, and Gemini CLI, plus similar AI coding assistants. It does not document native integrations for ChatGPT, Claude AI, the OpenAI API, or the Claude API.
Can it test an application running locally or on a private network?
Yes, through the TestMu AI Tunnel. The appropriate Windows, macOS, or Linux tunnel must be configured, and the test capabilities must include tunnel: true.
Does this repository include the Rook skill for testing AI agents?
Not through the same installation path. Rook's skill source and documentation live in the separate LambdaTest/rook repository, and its installer is npx @testmuai/rook-skill; it requires Node.js 22 or newer.
What does cloud testing cost?
Pricing, plans, and free-tier limits are not specified in the supplied material, so those commercial terms need to be confirmed before adoption.
View on GitHub ↗ Install ↓

Compare agents like this one

The same FARS review applied across the shortlist this agent qualifies for.

Related agents