lettuce-skill — Claude Code skill: what it does, how to install
Generates Python BDD tests using the legacy Lettuce framework. Builds Gherkin feature files, step definitions, and terrain hooks for existing test suites.
MIT 208 installs ★ 393 updated 2026-10-13
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a claude-codeWhat lettuce-skill does
This skill outputs Gherkin .feature files, Selenium step definitions using Lettuce decorators, and terrain.py lifecycle hooks. It handles local test execution and remote runs on LambdaTest using LT_USERNAME and LT_ACCESS_KEY credentials.
Lettuce is an unmaintained legacy library, so new Python projects should use Behave instead. This skill exists strictly to maintain older codebases where Lettuce is already entrenched. Running the generated code requires installing lettuce and selenium via pip.
What it helps with
- write Gherkin feature files for legacy Lettuce suites
- generate Python step definitions using
@stepdecorators and Selenium - configure setup and teardown lifecycle hooks inside
terrain.py - configure remote cloud browser execution for LambdaTest
Example prompts for the agent
- Generate a Lettuce login feature with matching steps.py definitions
- Create a terrain.py file that closes the Selenium browser after each Lettuce scenario
- Add step definitions for checking page text in our lettuce BDD tests
How to install lettuce-skill
1Skills CLI (Vercel)
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a claude-code2Manually: put the files in place
~/.claude/skills/lettuce-skill/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/lettuce-skill/
1Skills CLI (Vercel)
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a codex2Manually: put the files in place
.agents/skills/lettuce-skill/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a cursor2Manually: put the files in place
.cursor/skills/lettuce-skill/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install LambdaTest/agent-skills lettuce-skill2Manually: put the files in place
.github/skills/lettuce-skill/SKILL.md1Skills CLI (Vercel)
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a gemini-cli2Platform CLI
gemini skills install https://github.com/LambdaTest/agent-skills --path lettuce-skillManually: put the files in place
.devin/skills/lettuce-skill/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add LambdaTest/agent-skills --skill lettuce-skill -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about lettuce-skill
How to install lettuce-skill in Claude Code?
Run npx skills add LambdaTest/agent-skills --skill lettuce-skill -a claude-code in your terminal. You can also place the skill file directly at ~/.claude/skills/lettuce-skill/SKILL.md.
What does lettuce-skill do?
It prompts the assistant to produce Lettuce BDD files, including Gherkin feature specifications, Python step functions, and setup hooks.
Should I use Lettuce for a new Python project?
No. Lettuce is no longer maintained, so Behave is the recommended choice for modern Python BDD suites.
Source text
SKILL.md · MIT
---
name: lettuce-skill
description: >
Generates Lettuce BDD tests for Python with feature files and step definitions.
Note: Lettuce is legacy/unmaintained; consider Behave for new projects. Use when
user specifically mentions "Lettuce". Triggers on: "Lettuce", "lettuce test",
"lettuce BDD".
languages:
- Python
category: bdd-testing
license: MIT
metadata:
author: TestMu AI
version: "1.0"
---
# Lettuce BDD Skill (Legacy)
> **Note:** Lettuce is largely unmaintained. For new Python BDD projects, use **Behave** instead.
## Core Patterns
### Feature File (features/login.feature)
```gherkin
Feature: User Login
Scenario: Successful login
Given I navigate to the login page
When I enter "[email protected]" as email
And I enter "password123" as password
And I click login
Then I should see the dashboard
Scenario: Invalid login
Given I navigate to the login page
When I enter "[email protected]" as email
And I enter "wrong" as password
And I click login
Then I should see "Invalid credentials"
```
### Step Definitions (features/steps.py)
```python
from lettuce import step, world
from selenium import webdriver
from selenium.webdriver.common.by import By
@step(r'I navigate to the login page')
def navigate_to_login(step):
world.browser = webdriver.Chrome()
world.browser.get(world.base_url + '/login')
@step(r'I enter "([^"]*)" as email')
def enter_email(step, email):
el = world.browser.find_element(By.ID, 'email')
el.clear()
el.send_keys(email)
@step(r'I enter "([^"]*)" as password')
def enter_password(step, password):
el = world.browser.find_element(By.ID, 'password')
el.clear()
el.send_keys(password)
@step(r'I click login')
def click_login(step):
world.browser.find_element(By.CSS_SELECTOR, 'button[type="submit"]').click()
@step(r'I should see the dashboard')
def see_dashboard(step):
assert '/dashboard' in world.browser.current_url
@step(r'I should see "([^"]*)"')
def see_text(step, text):
assert text in world.browser.page_source
```
### Terrain (Setup/Teardown — terrain.py)
```python
from lettuce import before, after, world
@before.all
def setup():
world.base_url = 'http://localhost:3000'
@before.each_scenario
def before_scenario(scenario):
pass
@after.each_scenario
def cleanup(scenario):
if hasattr(world, 'browser'):
world.browser.quit()
@after.all
def teardown(total):
print(f"Ran {total.scenarios_ran} scenarios")
```
## Setup: `pip install lettuce selenium`
## Run: `lettuce` or `lettuce features/login.feature`
### Cloud Execution on TestMu AI
Set environment variables: `LT_USERNAME`, `LT_ACCESS_KEY`
```python
# terrain.py
from selenium import webdriver
import os
@before.all
def setup():
lt_options = {
"user": os.environ["LT_USERNAME"],
"accessKey": os.environ["LT_ACCESS_KEY"],
"build": "Lettuce Build",
"platformName": "Windows 11",
"video": True,
"console": True,
}
options = webdriver.ChromeOptions()
options.set_capability("LT:Options", lt_options)
world.browser = webdriver.Remote(
command_executor=f"https://{os.environ['LT_USERNAME']}:{os.environ['LT_ACCESS_KEY']}@hub.lambdatest.com/wd/hub",
options=options,
)
```
## Migration: Consider switching to `behave` for active maintenance
## Deep Patterns
For advanced patterns, debugging guides, CI/CD integration, and best practices,
see `reference/playbook.md`.
© TestMu AI · MIT · LambdaTest/agent-skills