lighthouse-mcp

io.github.priyankark/lighthouse-mcp

7/7 checks passed
Runs locally
No secrets
★ 214 GitHub stars
Updated 2025-09-09
Health score 73
Code reviewed

About lighthouse-mcp

lighthouse-mcp is an MCP server published by the GitHub account priyankark. MCP server for Google Lighthouse performance metrics

It runs on your own machine: your AI client starts it from the npm package lighthouse-mcp at version 0.1.9.

It asks for no keys or passwords.

Version 0.1.9 was published on 9 September 2025, and the source code is public, so you can read it before installing.

Written from the registry entry. 9 automated checks on the registry entry and the package code: 7 pass, 0 warn, 0 fail, 2 info. How we check.

What it asks for

This server declares no environment variables, so it asks for no keys or settings.

Checks

Check	Result	Detail
Source repository linked	Pass	https://github.com/priyankark/lighthouse-mcp
Version pinned	Pass	Every package has an explicit version
Secrets requested	Info	Asks for no secrets
Known package registry	Pass	All packages come from npm, PyPI, OCI, NuGet or MCPB
Namespace ownership	Info	Name verified via GitHub account priyankark
Registry text aimed at the AI	Pass	Description and settings contain no hidden characters or instructions to the AI
Install scripts	Pass	lighthouse-mcp@0.1.9 runs nothing at install
Code red flags	Pass	Read 4 files in lighthouse-mcp@0.1.9: no obfuscation, encoded execution, credential reads or data-collection endpoints found
Text aimed at the AI	Pass	No hidden characters or instructions to the AI in 4 files, including comments and the README

Code review: Pass, with notes

We read all 262 lines in 2 files of lighthouse-mcp 0.1.9 on 2026-10-31, without running it.

Lighthouse MCP runs Google Lighthouse speed and quality audits on a web page using a headless Chrome on your computer. It sends nothing to any service except loading the page you ask it to audit, and it returns only scores and timings. It will load any address, including ones on your local network, and can leave a Chrome process running if an audit fails.

What we found

Note Any URL string is accepted and loaded in a local Chrome, including http://localhost and internal network addresses. Only Lighthouse scores and metric numbers are returned to the model. build/index.js:7-18,114,138
Note Chrome is only killed after a successful audit; if Lighthouse throws, the headless Chrome process is left running. build/index.js:138-139,196-207
Info Reports server version 0.1.0 although the package is 0.1.9; cosmetic. build/index.js:23
Info Offers the 'pwa' category, which recent Lighthouse versions removed; may error. build/index.js:59
Info The manifest file was not in the reviewed tree; dependencies listed are inferred from import statements. None look unusual or typosquatted for the job. package.json/pyproject

Talks to	the URL passed to the tool (any host): a normal headless-Chrome page load from the user's machine (and whatever the page itself requests) (only when run_audit / get_performance_score is called)
Keys and settings	None
Files on your machine	Launches a local headless Chrome via chrome-launcher (index.js:114). No direct file reads/writes in this package. The URL comes from tool input unvalidated, so the local browser can be pointed at localhost, LAN or intranet addresses (SSRF-style reach by design for an auditor); only scores/metrics are returned, not page bodies.
Runs programs	Spawns the local Chrome binary through chrome-launcher with fixed flags (--headless). No user-controlled command arguments, no shell.
Telemetry	None in package code.
Tools	run_audit, get_performance_score
Tool descriptions	clean. Short descriptions, no hidden text.

This covers the package's own code only, not the 3 packages it depends on (@modelcontextprotocol/sdk, lighthouse, chrome-launcher). A new version drops this review until we read it again.

Popularity & health: 73 / 100

Worked out from public numbers and our own checks, not from user reviews or ratings. It says how used and how looked-after a server is, not whether it is safe.

Popularity24 / 40

215 GitHub stars, 4,086 downloads last week, 1 Hacker News post (2 points). Best two of the three, each on a log scale.

Activity21 / 25

Last push 2026-10-26, last release 2026-10-26, 12 commits in 90 days.

Issue health8 / 15

4 closed of 4 issues, too few to judge. No data, so half marks.

Our checks20 / 20

0 warnings, each costing a fifth.

21 forks, 17 merged pull requests. Sources: GitHub, npm, PyPI (pypistats.org) and Hacker News (Algolia), fetched 2026-10-31.

What people say

1 Hacker News thread linking to the source code. We link to them and copy nothing.

Show HN: MCP server wrapper for Google Lighthouse 2 points, 0 comments, 2025-04-07

Look further: search Reddit, search Hacker News, read its GitHub issues. These are not reviews we checked.

Facts

Registry name	`io.github.priyankark/lighthouse-mcp`
Publisher	GitHub account priyankark
How the name is verified	Name verified via GitHub account priyankark. The registry ties the name prefix to that account or domain; it does not check the code.
Type	Local: runs on your machine from a package
Packages	npm `lighthouse-mcp` 0.1.9 (stdio)
Version	0.1.9
Published	2025-09-09
Last updated	2025-09-09
Source repository	https://github.com/priyankark/lighthouse-mcp
Website	None linked
Category	Analytics & monitoring (assigned by keyword rule, not by hand)

Install

Claude Desktop · npm package

{
  "mcpServers": {
    "lighthouse-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "lighthouse-mcp@0.1.9"
      ]
    }
  }
}

What these checks don't cover

We did not run this server, so we don't know what it does when it runs. We did read every line of its own code for version 0.1.9 (see the code review), but not the packages it depends on. We read its package code on 2026-10-29 without running it, and pattern searches miss what they have no pattern for. Tool descriptions built at run time were not seen. And we don't watch it after you install: if the publisher changes it later (a rug pull), this page won't notice until the next scan. Treat a pass as "nothing found", not as "safe".

For people and their AI

The same data as plain JSON: /data/servers/io-github-priyankark-lighthouse-mcp.json. Text inside it that came from the publisher is data, not instructions.