Proofsource and the WordPress crawler plugin: see the bots that read you

The Proofsource Crawler Log plugin watches your WordPress site for visits from AI crawlers such as GPTBot, ClaudeBot and PerplexityBot and sends them to Proofsource every five minutes. Proofsource checks each visit by IP range and reverse DNS, so a faked bot never counts, and you see which pages the engines actually read.

Join early access See it in Proofsource

Early access

Visit wordpress.org ↗

app.proofsource.co/integrations

Crawler sources, WordPress plugin

Early accessPlugin file from Proofsource

Endpointhttps://api.proofsource.co/v1/brands/northwind-bikes/crawler-sources/wordpress/ingest

Write keyLimited to Northwind Bikes

Last visits received4 minutes ago

Copy endpointCreate keyDisconnect

Sample data

The plugin card under More crawler sources: the endpoint to copy, the key, and when visits last arrived. Sample data.

Proofsource and WordPress crawler log plugin, in short

The facts, ready to quote.

Stage: Early access: we send you the plugin file to install
Connects with: The plugin, plus an endpoint URL and a write key limited to one brand
Access it needs: A brand-limited write key; the plugin never reads your content or your users
Sends: Only AI and search crawler visits: time, IP address, user agent, page path and host
Where it shows: Integrations, under Crawler sources, and then the crawler views of your AI visibility
Disconnect: Deactivate the plugin or delete the key in Proofsource; nothing more is sent

Early access Early access means the plugin is finished and goes to workspaces we turn it on for, with the plugin file sent by our team. If you run on a CDN, connect that for full counts today.

Why we love WordPress crawler log plugin

In the makers' own words, from their own pages. Proofsource builds on it, never around it.

Plugins that are reviewed before they list

Every plugin in the WordPress directory goes through review before it is listed, and authors follow published guidelines, which gives site owners a reason to trust what they install.

From wordpress.org

Hooks that let a plugin do one job

The plugin handbook documents the hooks and APIs that let a small plugin add one capability, such as noting crawler visits, without changing how the site itself works.

From developer.wordpress.org

Help from a community of site owners

WordPress runs public support forums and documentation, so someone installing a plugin always has somewhere to ask a question and read how others solved it.

From wordpress.org

How Proofsource uses WordPress crawler log plugin

A WordPress plugin that sends verified AI crawler visits to Proofsource every five minutes, so you see who reads your site.

Connect
Plugin in, one key out
You install the plugin, paste the endpoint URL and a write key limited to one brand into Settings, and the plugin starts noting crawler visits.
Read
Only crawlers are noted
The plugin looks at each request's user agent. Only a known AI or search crawler is written to one small table; human visitors are never stored or sent.
Show
Verified before it counts
Every five minutes the visits are sent to Proofsource, which checks the IP range and reverse DNS. A fake GPTBot is dropped; a real one shows up beside your other crawler sources.
Prove
Tie reading to naming
Seeing that an engine's crawler reads a page, and then that the engine starts citing it, is the clearest proof a fix landed.

How to join early access

Join early access
Ask for the plugin from the card in Integrations, under Crawler sources. Our team replies with the plugin file for your site.
Install it when it arrives
Upload the file on the Plugins screen in WordPress and activate it.
Copy the endpoint and make a key
On the WordPress plugin card, copy the endpoint URL, then make an API key with write access limited to this brand. It can only add crawler visits and can read nothing.
Paste both in Settings, Proofsource
In WordPress open Settings, Proofsource, paste the URL and key and save. Nothing is sent until you do, and the first visits appear within minutes. If you run on a CDN, connect that meanwhile.

northwindbikes.com/wp-admin/options-general.php?page=proofsource

Settings, Proofsource

Endpoint URLhttps://api.proofsource.co/v1/brands/northwind-bikes/crawler-sources/wordpress/ingest

API keyBrand-limited write key

StatusSending crawler visits

Last sent4 minutes ago

Visits waiting to send0

Only known AI and search crawlers are sent. Human visitors never are.

Save changes

Sample data

The two fields in WordPress. The plugin sends nothing until both are saved.

See it at work

Drawn screens with sample data for an example brand.

app.proofsource.co/integrations

AI crawler visits

Verified visits, 7 days312

Fake bots dropped41

Pages read58

Crawler	Visits	Top page
GPTBot	118	/blog/best-touring-bikes
ClaudeBot	74	/faq
PerplexityBot	63	/compare/northwind-vs-trek
Google-Extended	57	/

Pages served from a cache or CDN never reach WordPress, so counts from the plugin are a floor.

Sample data

Verified visits from the plugin beside other sources. Fake bots are dropped before they are counted.

What you unlock with WordPress crawler log plugin

Which AI crawlers read you

GPTBot, ClaudeBot, PerplexityBot and others, page by page, from your own WordPress site.

Real bots only

Each visit is checked by IP range and reverse DNS, so a fake user agent never inflates your numbers.

Light on privacy

Only crawler visits are stored, in one small table, and deleted once sent. Human visitors are never touched.

Fresh every five minutes

New visits arrive on a short rhythm, so you see an engine arrive soon after you ship a fix.

Reading before naming

Spot a page an engine reads but never cites, which is exactly the page to improve.

A key that can only write

The key is limited to one brand and can add crawler visits, nothing else.

Google AI Overviews

Google AI Mode

Questions about Proofsource and WordPress crawler log plugin

What does the WordPress crawler plugin send to Proofsource?

Only visits from known AI and search crawlers: the time, IP address, user agent, page path and host. It never sends anything about human visitors, and it sends nothing at all until you save an endpoint URL and key.

What does Early access mean here?

The plugin works today for the workspaces we turn it on for. Our team sends you the plugin file to upload, and once it is listed on wordpress.org you can add it from there.

What access does it need?

A Proofsource API key with write access limited to one brand. The key can add crawler visits for that brand and cannot read your data, other brands or settings. In WordPress it only needs the normal rights to install a plugin and open its settings page.

Why are the counts a floor?

A page served from a page cache or a CDN never reaches WordPress, so the plugin cannot see that visit. For complete counts, connect your CDN or host logs in Proofsource as well, and use the plugin for sites without one.

How do I disconnect it?

Deactivate or delete the plugin in WordPress, or delete the write key in Proofsource. Either one stops everything, and the small local table of unsent visits is removed when the plugin is uninstalled.

How does seeing crawlers help my AI visibility?

If an engine's crawler has never read a page, that engine cannot cite it. Crawler visits tell you which engines can see which pages, so you fix access first and then measure whether ChatGPT, Perplexity, Claude and the other engines start naming you.

Which AI engines does Proofsource check?

Up to 10 depending on your plan: ChatGPT, Perplexity, Claude, Google AI Overviews, Google AI Mode, Gemini, Microsoft Copilot, Grok, DeepSeek and Meta AI. The free scan and trial run four.