voxelum/x-minecraft-launcher/chrome-devtools-cli

Trust tier 2
  • community
LLM review pending
Source check pending

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

Use when: validating XMCL UI, Electron main/preload/runtime integration, renderer behavior, console errors, network requests, screenshots, or performance in the real development launcher through the Chrome DevTools CLI.

Install

npx vskill@latest install voxelum/x-minecraft-launcher/chrome-devtools-cli

SKILL.md

Chrome DevTools CLI

Use the official chrome-devtools CLI to inspect and drive the running XMCL Electron app. Do not configure, enable, or call Chrome DevTools through MCP tools. The npm package name still contains mcp, but the chrome-devtools binary is its standalone CLI client and does not require an MCP client config.

Use this command prefix without adding a project dependency:

bash
npx -y -p chrome-devtools-mcp@latest chrome-devtools

Connect to XMCL

  1. Reuse the running dev:main and dev:renderer tasks. Do not start duplicate

tasks or another Electron app.

  1. Discover the active debugging port from the Electron process arguments or

task output. Probe likely ports only as a fallback:

``bash curl --fail --silent http://127.0.0.1:9222/json/list curl --fail --silent http://127.0.0.1:9300/json/list ``

  1. Stop a stale CLI daemon, connect it to the discovered port, then verify the

selected launcher page:

``bash npx -y -p chrome-devtools-mcp@latest chrome-devtools stop npx -y -p chrome-devtools-mcp@latest chrome-devtools start --browserUrl http://127.0.0.1:<port> npx -y -p chrome-devtools-mcp@latest chrome-devtools list_pages --output-format=json ``

The real development page must be titled X Minecraft Launcher or XMCL and load from http://localhost:3000, not http://xmcl.runtime. Keep its numeric page ID for every page-scoped command.

Drive and inspect

Take a fresh accessibility snapshot before resolving element UIDs. UIDs can change after navigation, hot reload, dialogs, or other DOM updates.

bash
npx -y -p chrome-devtools-mcp@latest chrome-devtools take_snapshot <pageId> --output-format=json
npx -y -p chrome-devtools-mcp@latest chrome-devtools click <pageId> <uid>
npx -y -p chrome-devtools-mcp@latest chrome-devtools fill <pageId> <uid> "value"
npx -y -p chrome-devtools-mcp@latest chrome-devtools press_key <pageId> "Enter"
npx -y -p chrome-devtools-mcp@latest chrome-devtools evaluate_script "() => ({ title: document.title, href: location.href })" --pageId <pageId> --output-format=json

Inspect renderer failures and relevant requests after exercising the complete workflow:

bash
npx -y -p chrome-devtools-mcp@latest chrome-devtools list_console_messages <pageId> --output-format=json
npx -y -p chrome-devtools-mcp@latest chrome-devtools list_network_requests <pageId> --output-format=json

Use <command> --help before relying on an unfamiliar or version-sensitive flag. Prefer JSON output so results can be checked precisely. Screenshots for temporary inspection belong outside the repository unless the task explicitly requires review evidence:

bash
npx -y -p chrome-devtools-mcp@latest chrome-devtools take_screenshot <pageId> --filePath "$TEMP/xmcl-devtools.png"

For the main launcher, validate at 1200x720 and, when layout risk warrants it, 800x400. Do not use mobile-like viewport sizes.

Finish

  • Preserve user state and remove only temporary resources created by the test.
  • After a main/runtime rebuild, restart only this repository's Electron app,

reconnect the CLI, and repeat the same workflow.

  • Run the narrowest automated test plus type-check or lint required by

AGENTS.md; CLI validation does not replace those checks.

  • Stop the CLI daemon when validation is complete:

``bash npx -y -p chrome-devtools-mcp@latest chrome-devtools stop ``

  • Report the route and controls exercised, observed state transitions, relevant

console/network results, screenshots if requested, and cleanup performed.

Security

  • Security scanPassed, 93 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

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

Source

Popularity

Repository stars
1,609
Repository forks
204
7-day trend score
37

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 24 Oct 2026, 4 days ago.

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/voxelum/x-minecraft-launcher/chrome-devtools-cli/badge)](https://verified-skill.com/skills/voxelum/x-minecraft-launcher/chrome-devtools-cli)

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