How to Run Espresso Tests With Agent Skills on TestMu AI
Running Espresso tests with Agent Skills on TestMu AI lets your AI coding assistant generate Android UI automation from natural-language prompts, scaffolding Gradle configuration, test code, and cloud execution so you run Espresso suites on real devices without manual setup.
Instead of manually wiring up your Espresso project, use Agent Skills to let your AI coding assistant generate production-ready Android UI automation for you. The espresso-skill is part of TestMu AI Agent Skills - structured packages that teach AI assistants how to write production-grade test code and run it on real devices.
What Agent Skills Are
Self-contained packages of instructions, code patterns, and configurations for a specific testing framework.
Once installed, your AI assistant works like a senior Android QA automation architect with knowledge of:
- Proper Espresso project structure in Java and Kotlin
- Correct Gradle dependencies and build configuration
- Both local and TestMu AI real-device cloud execution patterns
- Common pitfalls and debugging approaches
- CI/CD integration with GitHub Actions
Prerequisites
Make sure you have the following set up before you start.
- Create a TestMu AI account if you don't have one.
- Get your Username and Access Key from the TestMu AI Dashboard.
- Install an AI coding assistant: Claude Code, GitHub Copilot, Cursor, Gemini CLI, or any compatible tool.
Step 1: Install the Espresso Agent Skill
Clone the repository and copy the Espresso skill into your AI tool's skills directory.
git clone https://github.com/LambdaTest/agent-skills.git
# For Claude Code
cp -r agent-skills/espresso-skill .claude/skills/
# For Cursor
cp -r agent-skills/espresso-skill .cursor/skills/
# For GitHub Copilot
cp -r agent-skills/espresso-skill .github/skills/
# For Gemini CLI
cp -r agent-skills/espresso-skill .gemini/skills/
Install all available framework skills at once by cloning the repository directly into your tool's skills directory (e.g., .claude/skills/, .cursor/skills/).
Step 2: Set Your Credentials
Add your TestMu AI credentials as environment variables so the generated tests can authenticate with the real device cloud.
Visit the TestMu AI Dashboard, navigate to the left sidebar, and select Credentials. Copy your Username and Access Key, then run:
export LT_USERNAME="YOUR_LAMBDATEST_USERNAME"
export LT_ACCESS_KEY="YOUR_LAMBDATEST_ACCESS_KEY"
Step 3: Run Your First Test
Ask your AI assistant to write tests in natural language.
"Write an Espresso test for my Android app's login screen and run it on TestMu AI real devices"
"Set up an Espresso project in Kotlin with a RecyclerView test and run it on TestMu AI"
"Run my Espresso test suite across a Pixel 7 and a Galaxy S23 in parallel on TestMu AI, tag the build as 'release-v1.0'"
The Agent Skill automatically handles:
- Project setup - Creates the right Gradle configuration, dependencies, and instrumentation runner
- Language detection - Maps your request to Java or Kotlin
- Execution target - Configures local or TestMu AI real-device execution based on your request
- Best practices - Uses Espresso idioms (ViewMatchers, ViewActions, ViewAssertions) and IdlingResource for synchronization
What's Included in the Espresso Skill
A breakdown of everything the skill package contains.
espresso-skill/
├── SKILL.md
└── reference/
├── playbook.md
└── advanced-patterns.md
| Section | Details |
|---|---|
| Project Setup | Gradle dependencies, instrumentation runner, project structure |
| Core Patterns | Complete, runnable code examples with best practices |
| Cloud Integration | TestMu AI real-device capabilities and app-upload configuration |
| CI/CD Integration | GitHub Actions workflows with reporting and parallel execution |
| Debugging Guide | Common build and shard errors mapped to solutions |
| Best Practices | Actionable recommendations for production Android test code |
Supported AI Tools
Pick the tool you use and install the skill to the corresponding path.
| Tool | Type | Installation Path |
|---|---|---|
| Claude Code | CLI | .claude/skills/ |
| GitHub Copilot | Extension | .github/skills/ |
| Cursor | IDE | .cursor/skills/ |
| Gemini CLI | CLI | .gemini/skills/ |
| Codex CLI | CLI | .codex/skills/ |
| OpenCode | CLI | .opencode/skills/ |
| Claude.ai | Web | Settings > Features > Skills |
Next Steps
Continue with these related guides:
