Raise a WordPress site's PageSpeed score with Cursor

Raising a WordPress PageSpeed score with Cursor means asking its agent to preview and then apply an xSpeed optimization pass through xSpeed Hub, while the theme code stays open beside the results.

list_sitesget_score_historylist_modulesoptimize_sitegenerate_critical_csspurge_cacherun_speed_test

Last updated: October 2026

Example session · Cursor

You: With xSpeed Hub, raise the PageSpeed score on the agency site. Preview first.

list_sites read-only
→ 5 sites; agency is agency.example.com

optimize_site(site: agency, aggressiveness: standard, dry_run: true) dry run
→ plan: 5 settings to apply (CSS minify, defer JS, lazy-load images, compression, browser cache), nothing changed

Cursor: The plan is CSS minify, script deferral, lazy-loading, compression and browser caching. Nothing has changed. Apply it to agency?

You: Yes, apply it.

optimize_site(site: agency, aggressiveness: standard) changes site
→ applied 5, undone 0, unfixable: images without width and height, 1 hotlinked script

Cursor: All five settings are on. Two items are left that caching cannot fix: images missing width and height, and one script loaded from another domain. If both come from this theme, I can open the template files and propose edits. First, open the home page in a browser and check the console is clean.

How Cursor handles this

Cursor is the place this job gets interesting, because the theme is already in the editor. You ask the agent to raise the score on a site, and it calls list_sites and then optimize_site with dry_run on. What happens next depends on your Run Mode: under Allowlist the call waits for your approval and you can expand it to see its arguments, while under the default Auto-review a classifier handles calls that are not listed. Then you see the plan: the settings xSpeed would turn on, in order, with nothing changed on the site yet.

Approve the real pass when the plan reads right. xSpeed applies each setting one at a time, checks the page after each, and undoes anything that breaks it, taking up to two minutes per site. The result splits cleanly in two. The applied and undone lists describe what the plugin did. The unfixable list describes what caching cannot touch: page weight, images hotlinked from another domain, a heavy video, DOM size. Those are code and content problems, and they are exactly what you can open in the editor and fix.

Ask the agent to turn that list into edits in your theme, such as a missing width and height on images or a script loaded in the header that could move to the footer. Then run the pass again. A site that is already tuned returns that everything is already on, which is a finished answer. Last, the Hub gives the agent verify_urls and tells it to look at the pages before claiming success. Cursor cannot see a browser console for you, so open them yourself.

Set up Cursor once

Already connected? Skip to the prompts. Alternatives and troubleshooting are on the Cursor guide.

1Put your sites in xSpeed Hub

Sign in at app.xspeedcache.com with Google or email; the Hub is free and has no site cap. Then connect each WordPress site from its own dashboard: click Connect Hub in the xSpeed Cache top bar, then Connect via xSpeed Hub. Each site needs the free xSpeed Cache plugin.

2Add xSpeed Hub to mcp.json

Add the server to ~/.cursor/mcp.json to have it in every project, or to .cursor/mcp.json inside a project to scope it there. The entry is just a name and a url.

~/.cursor/mcp.json

{
  "mcpServers": {
    "xspeedhub": {
      "url": "https://app.xspeedcache.com/xspeed/mcp"
    }
  }
}

3Enable it and sign in

Open Customize in the sidebar, find xspeedhub and make sure it is switched on. When Cursor starts the OAuth sign-in, approve access on the xSpeed Hub page. There is no token to paste. If the server shows an error, open the Output panel with Cmd+Shift+U and pick MCP Logs.

Text

Customize  >  xspeedhub  (on)

1 / 3

Full Cursor setup, sign-in options and FAQ

Before you send a prompt that changes something

Reads change nothing on your sites, though contact_support emails xSpeed support. Writes run once your connection allows them, so your client's approval prompt and a read-only connection (the token's Read-only everywhere switch, or a Viewer sign-in) are the gates that matter. Cursor decides how to treat an MCP tool call by its Run Mode, and you can expand a call to see its arguments. MCP follows the same Run Modes as terminal commands, set under Settings > Agents > Approvals & Execution. The recommended default, Auto-review, runs allowlisted calls at once and sends the rest to a classifier, not to you. Allowlist runs only what you listed, and with an empty list it asks you every time. Run Everything runs every tool call without asking. xSpeed Hub has no confirmation step of its own, so a tool you allowlist, or any call under Run Everything, writes as soon as your connection allows writes. Auto-review is a model, and Cursor says it is not a security boundary.

What do I ask?

Three prompts written for Cursor. More for this job are below.

Prompt for Cursor

Use xSpeed Hub to preview an optimization for the agency site, standard level. Do not apply anything yet.

Prompt for Cursor

Apply the plan with xSpeed Hub, then turn the unfixable list into concrete edits in this theme and show me the diff before saving.

Prompt for Cursor

With xSpeed Hub, run the optimization again on the agency site and tell me whether it now says everything is already on.

What happens, step by step

optimize_site is the Hub tool for "make my site faster". It measures the site, applies the recommended xSpeed settings one at a time, checks the page still renders after each change, undoes anything that breaks it, and reports what it applied, what it undid and what it could not fix. It reaches caching and asset delivery only. Page weight, images hosted on another domain and heavy video are outside it, and the report says so instead of claiming a win.

See where the site stands

The agent reads get_score_history for the last stored scores and list_modules for which modules exist on that site and what state they are in. Neither changes anything.

Preview with dry_run

optimize_site with dry_run: true returns the plan: each change and its tier, what would be skipped (already on, or needing a higher aggressiveness) and what it cannot fix. Nothing is applied. The tool description tells the agent to prefer this before the first real run on a site you have not optimized before.

Apply it

optimize_site without dry_run applies the plan. aggressiveness is safe (removals and server-side changes only), standard (the default) or aggressive, which includes settings known to break some themes and should be opted into deliberately. A run can take up to about 2 minutes per site.

Look at the site afterwards

When changes were applied, the result carries verify_urls and a note. The site checks its own HTML and reverts anything that breaks it, but it cannot run JavaScript, so verified: true does not prove the page works in a browser. The Hub tells the agent to open those URLs and check the page renders and the console is clean, or to tell you which URLs to check.

Reach for a number only if you named one

target_score (1 to 100, Pro sites only) turns one pass into repeated rounds toward that score, up to max_rounds, with 3 as the most per site. Each round is a real PageSpeed measurement and up to about 2 minutes. The result says why it stopped, and diminishing returns means what is left is outside what caching can reach.

Critical CSS on Pro sites

generate_critical_css generates above-the-fold CSS so the first screen can paint without waiting for the full stylesheets. It needs the Pro Critical CSS module on that site, and the site answers with an unknown-tool error without it. Pages are rebuilt with it on their next build, so the agent purges the cache afterwards.

Reference

Main tool	optimize_site (write): measure, apply one setting at a time, verify, undo what breaks, re-measure
Preview	dry_run: true returns the plan and changes nothing
aggressiveness	safe, standard (default) or aggressive
Every site at once	site: "all" or a list of handles, one result per site, sites run one after another; the Hub's confirm-once instruction does not name optimize_site, so your client's prompt is the gate
Time	Up to about 2 minutes per site
target_score	1 to 100; needs Pro on the site; without it the site runs one pass and says so
max_rounds	Up to 3 per site; only sent along with target_score; 12 rounds in total per call
Wide fan-out with a target	Each site gets fewer rounds, and the result carries a note saying so
Result fields	applied, skipped, reverted, unfixable, score, verified, rounds; verify_urls after a run that changed something
Already tuned site	Returns "everything is already on" plus what it cannot fix; that is an answer, not a failure
Critical CSS	generate_critical_css (write), Pro Critical CSS module required; purge_cache afterwards
Read-only connection	optimize_site and generate_critical_css are refused; dry_run is still a write call, so it is refused too. Applies to the connection token with Read-only everywhere on and to a Viewer sign-in, not to other OAuth sign-ins

Rules worth keeping

Run dry_run first. optimize_site changes live site settings as soon as your connection allows writes. The tool description tells the agent to preview first on a site that has not been optimized, but that is guidance, so your client's approval prompt and a read-only connection (the connection token with Read-only everywhere on, or a Viewer sign-in) are the real gates.
Keep aggressive for a deliberate choice. It includes settings known to break some themes, and the agent should say so before using it.
After a run that changed something, open the pages in verify_urls. The HTML checks cannot run JavaScript, so a page can pass them and still be broken in a browser. Do not call a run safe because verified is true.
Relay the unfixable list. When a site is already tuned, the honest answer is that caching is done and the rest is page weight, hotlinked images or DOM size. The agent should not retry the tool or claim a gain it did not make.
Only ask for target_score when you want the site driven to a number. Each round costs a real PageSpeed measurement, and a plain "make it faster" is one pass.

Good to know with Cursor

Cursor does not ask by default. Since Cursor 3.6 the recommended Run Mode is Auto-review, which sends calls that are not on your allowlist to a classifier that Cursor says is not a security boundary. Run Everything asks nothing, and only Allowlist mode asks you about calls that are not listed. The Hub has no confirmation step of its own, so under Run Everything, or once optimize_site is on your allowlist, it runs as soon as your connection allows writes. The preview and the real pass use the same tool, so one allowlist entry covers both. Allowlist the reads, keep optimize_site off the list while you tune a live site, and do not give a Cloud Agent a write-capable connection, because Cursor says Cloud Agents never ask.

More prompts for this job

They work in any client connected to xSpeed Hub.

Prompt

Ask xSpeed Hub to preview what optimize_site would change on shop with a safe pass. Do not apply anything.

Prompt

Use xSpeed Hub to optimize the blog site with the standard pass, then give me the pages to check by hand.

Prompt

With xSpeed Hub, get docs to a mobile PageSpeed score of 90 if you can, and tell me why it stopped if it does not get there.

Prompt

Use xSpeed Hub to show me what optimize_site would do on every site, as a dry run, and list which sites have the most left to change.

Prompt

Use xSpeed Hub to generate critical CSS for shop, then purge the cache so the pages are rebuilt with it.

Prompt

With xSpeed Hub, optimize shop, then compare the new PageSpeed score with the last stored one.

Prompt

Using xSpeed Hub, what could optimize_site not fix on the shop site?

Frequently asked questions

Allowlist is the most predictable. Switch Run Mode to Allowlist, list read tools such as list_sites, list_modules and get_score_history, and leave optimize_site off the list so each call asks. Under the default Auto-review, unlisted calls go to a classifier instead, and Cursor says it is not a security boundary. You change modes under Settings, Agents, Approvals and Execution.

It can edit the theme for items on the unfixable list, such as missing image dimensions or a script that loads from another domain. Review each diff yourself, deploy, and run the optimization again to see whether the list shrinks.

It measures the site, applies the recommended xSpeed settings one at a time, checks that the page still renders after each change, and undoes any change that breaks it. It then reports what was applied, skipped and reverted, and what it could not fix. It reaches caching and asset delivery only.

Yes, especially on a site you have not optimized before. dry_run: true returns the plan, what would be skipped and what cannot be fixed, and changes nothing. The tool description tells the agent to prefer it, but nothing forces it, so say so in your prompt if you want a preview.

Yes, with target_score, on a site that has xSpeed Pro. It repeats measure, apply and re-measure for up to 3 rounds per site and stops early when rounds stop helping, saying why. On a Free site the score target is ignored and you get a single pass with a note that no iterative tuner is installed.

That is a real answer. The settings it can apply were already enabled, so there was nothing to change. The result also lists what it could not fix, such as page weight, images hosted on another domain or DOM size, and those need changes outside a caching plugin.

The site checks its own HTML after each change and reverts anything that breaks it, but it cannot run JavaScript. After a run that changed something, the Hub tells the agent to open the returned verify_urls and check the page renders and the console is clean, or to give you the URLs to check yourself.

No. It needs the Pro Critical CSS module on that site, and without it the site answers with an unknown-tool error. On a Pro site, purge the cache afterwards, because pages already cached were built without the critical CSS.

Raise the PageSpeed score, from Cursor.

xSpeed Hub is free and has no site cap. Connect your sites once and ask from the agent you already use.

Open xSpeed Hub Cursor setup

Raise a WordPress site's PageSpeed score with Cursor

How Cursor handles this

Set up Cursor once

What do I ask?

What happens, step by step

Reference

Rules worth keeping

More prompts for this job

Frequently asked questions

Keep going

Raise the PageSpeed score, from Cursor.