juanmarchetto/agent-skills/e2e-pipeline

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Generate and run E2E tests for mobile and web apps. Discovers screens, states, and flows via static analysis, generates Maestro test flows with assertions, executes them, and reports pass/fail results. Use when: run e2e, test flows, e2e tests, generate tests, automated testing.

Install

npx vskill@latest install juanmarchetto/agent-skills/e2e-pipeline

SKILL.md

E2E Pipeline

Generate and run comprehensive E2E tests from static analysis of your codebase. Discovers all screens, states, and flows, generates Maestro test flows with real assertions, and reports results.

How This Differs From demo-pipeline

demo-pipeline produces marketing assets (video recordings, screenshots, animated demos). e2e-pipeline produces test results (pass/fail verdicts, regression detection, coverage reports). Same discovery phase, completely different output: demo-pipeline asks "does it look good?" while e2e-pipeline asks "does it work correctly?"

Workflow

Pipeline: DISCOVER > DETECT > GENERATE > EXECUTE > REPORT

Phase 0: DISCOVER — Flow & State Analysis

Analyze the codebase to map all screens, states, flows, and native dependencies.

Step 1: Discover screens Read the file-based routing structure:

bash
find src/app -name "*.tsx" -not -name "_layout.tsx" -not -name "+not-found.tsx" | sort

Step 2: Identify states per screen Search for UI conditionals:

bash
grep -n "loading\|isLogged\|isEmpty\|error\|balance.*>\|\.length.*==.*0" src/app/**/*.tsx

Step 3: Map navigation flows

bash
grep -rn "router\.push\|router\.replace\|href=" src/app/ src/components/

Step 4: Identify native dependencies

bash
grep -rn "expo-camera\|CameraView\|useMobileWallet\|SeedVault" src/

Step 5: Present flow map Present the discovered map and ask:

"I discovered N screens with M states and K flows. Which flows do you want to test? (default: all)"
Phase 1: DETECT — Project Type & State

Detect project type (Expo/React Native/web) and check running state (device/emulator/server).

Phase 2: GENERATE — Create Test Flows

For each selected flow, generate a Maestro test YAML with assertions.

Assertion Reference:

AssertionMaestro equivalentUse when
assert_visible: "Text"- assertVisible: "Text"Verify element appears on screen
assert_visible: { id: "elem" }- assertVisible: { id: "elem" }Verify element by testID
assert_not_visible: "Text"- assertNotVisible: "Text"Verify element is gone
assert_contains: { id: "elem", text: "val" }- assertVisible: { id: "elem", text: "val" }Verify element has specific text
assert_enabled: { id: "elem" }- assertVisible: { id: "elem", enabled: true }Verify element is interactive
assert_count: { text: "Item", count: 5 }- assertVisible: "Item" + repeat logicVerify expected number of elements
assert_text_contains: { id: "elem", text: "partial" }- assertVisible: { id: "elem", containsText: "partial" }Verify partial text match
screenshot_on_fail: trueMaestro auto-capturesCapture state on failure

Test YAML format:

yaml
flow: "<Flow Name>"
steps:
- name: "<Step Name>"
actions:
- tap: "Element"
- assert_visible: "Expected Result"
- screenshot_on_fail: true

Present for approval:

"I generated N test flows covering M screens. Review and approve, or select which to run."

HARD GATE: Do NOT execute until user approves.

Phase 3: EXECUTE — Run Tests

Step 1: Generate Maestro test files in ./maestro/tests/:

bash
mkdir -p maestro/tests

Step 2: Run all tests:

bash
maestro test maestro/tests/

Or run individually:

bash
maestro test maestro/tests/flow-name.yaml

Step 3: Capture failure screenshots:

bash
mkdir -p screenshots/failures
cp ~/.maestro/tests/*/screenshots/* screenshots/failures/ 2>/dev/null || true
Phase 4: REPORT — Results

Generate ./docs/test-report.md and present summary.

Example Output

E2E Test Results — 2026-03-16
Passed: 8/10 flows
Failed: 2 (see below)
FAILED checkout-flow:
assert_visible "Order Confirmed" timed out after 10s
Screenshot: failures/checkout-timeout.png
Likely cause: payment API not responding
FAILED search-flow:
assert_count items expected 5, got 0
Screenshot: failures/search-empty.png
Likely cause: API server not running or search index empty
Passed flows:
login-flow .............. OK (3.2s)
signup-flow ............. OK (4.1s)
profile-edit-flow ....... OK (2.8s)
settings-flow ........... OK (1.9s)
navigation-flow ......... OK (2.1s)
logout-flow ............. OK (1.5s)
onboarding-flow ......... OK (5.3s)
home-scroll-flow ........ OK (2.0s)
Coverage: 10 flows, 47 assertions, 32 screens
Report saved: ./docs/test-report.md

Error Handling

  • If Maestro is not installed: provide install command (curl -Ls "https://get.maestro.mobile.dev" | bash)
  • If no emulator/device detected: prompt user to start one and provide platform-specific instructions
  • If app is not running: attempt to start it, or ask user to launch manually
  • If test times out: capture screenshot, report which assertion failed, suggest increasing timeout
  • If all tests fail: likely environment issue — check device connection, app state, and API availability
  • If flow discovery finds 0 screens: ask user to specify the source directory

Structure

e2e-pipeline/
├── SKILL.md # This file
├── reference.md # Complete action and assertion reference
└── templates/
├── maestro-test.yaml # Maestro test template
└── test-report.md # Report template

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
5
Repository forks
1
7-day trend score
9

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/juanmarchetto/agent-skills/e2e-pipeline/badge)](https://verified-skill.com/skills/juanmarchetto/agent-skills/e2e-pipeline)

Something wrong with this skill or its page? Tell us.