Privacy Policy
Last updated: 11 August 2026 · Applies to the PagePull Chrome extension and pagepull.app
The short version. PagePull reads a web page only when you press Extract. There are two ways it can run the extraction, and you choose which in settings:
- Using your PagePull plan (the default) — the page's content is sent to our extraction service, which passes it to Anthropic and returns the table. We do not store the page content or your results; we keep only a count of how many extractions you have used, and the hostname of the page.
- Using your own Anthropic API key — the page's content goes from your browser straight to Anthropic and never reaches us at all.
Either way, your extracted tables stay on your computer, and we never sell your data or use it for advertising.
1. Who we are
PagePull ("the extension", "we") is a Chrome browser extension that extracts structured data from web pages at the user's request. For any privacy question, contact [email protected].
2. What the extension accesses
| Data | When | Where it goes |
|---|---|---|
| Web page content — the text and links of the page you are extracting | Only when you click Extract, run a saved recipe, or a monitor you created runs on its schedule | On a PagePull plan: to our extraction service at api.pagepull.app, which forwards it to the Anthropic API and returns the result. It is processed in memory to answer that one request and is not written to disk or retained.On your own API key: from your browser directly to Anthropic. It never reaches us. |
Usage record — a random installation ID, the number of extractions you have run, token totals, and the hostname of the page (for example example.com, never the full address or the page's contents) |
Each extraction, when you are on a PagePull plan | Stored by us, so we can enforce your monthly allowance and answer billing questions. The installation ID is a random identifier generated on your device; it is not linked to your identity unless you subscribe. |
| Network address | Each request to our service | Stored only as a truncated one-way hash, used to detect abuse of the free tier. The address itself is not retained. |
| Subscription details — your email address and subscription status | Only if you subscribe to a paid plan | Payment is handled by our reseller, Paddle, who acts as merchant of record. We never see or store your card details. We store your email address and subscription status in order to provide the plan and support it. |
| Your Anthropic API key (only if you choose that mode) | When you enter it in settings | Stored on your device only (chrome.storage.local). Used solely to authenticate your requests to Anthropic. Never synced to your Google account and never transmitted to us. |
| Your extraction settings and results — columns, saved recipes, monitors, extracted tables | As you use the extension | Stored on your device only. Never transmitted to us. |
3. What we do not do
- We do not retain the content of the pages you extract, or the tables produced from them.
- We do not collect analytics, usage telemetry, or crash reports.
- We do not track your browsing history. On a PagePull plan we record the hostname of pages you explicitly extract, and nothing else about your browsing.
- We do not sell, rent, or transfer your data to third parties.
- We do not use your data for advertising, credit assessment, or lending.
- We do not use your data to train any model. Anthropic does not train on API inputs or outputs either.
- We do not read pages in the background or on sites you have not asked us to work with.
- No human at PagePull reads your page content or results.
4. Third-party processing
Anthropic
To turn a page into a table, the page's simplified content is sent to the Anthropic API. On a PagePull plan this happens from our service using our own credentials; on your own API key it happens directly from your browser, using your key and under your own agreement with Anthropic. Anthropic's privacy policy and commercial terms are at anthropic.com/legal/privacy.
Before extracting from a page containing confidential or personal information, please make sure that sending it to Anthropic — and, on a PagePull plan, through our service — is consistent with your own obligations. If that matters to you, switch to your own API key in settings and the content will never reach us.
Paddle
If you subscribe, payment is processed by Paddle as merchant of record. They collect and hold your payment details; we never receive them. See paddle.com/legal/privacy.
Where our service runs
Our extraction service runs on servers we control in Nuremberg, Germany (Hetzner). Requests to it are encrypted in transit.
5. Permissions and why they are needed
- activeTab and scripting — to read the content of the tab you invoke the extension on, so it can be extracted.
- Optional site access (requested per site, only when you save a recipe) — so that saved recipes and monitors can re-run on that specific site without you re-authorising each time. It is requested one origin at a time, always as a result of an explicit action by you.
- storage — to keep your settings, recipes, monitors and results on your device.
- alarms — to run monitors on the schedule you chose.
- notifications — to alert you when a monitored page's data has changed.
- sidePanel — the extension's interface is a Chrome side panel.
- Host access to api.anthropic.com — to send extraction requests to the Anthropic API when you use your own key.
- Host access to api.pagepull.app — to send extraction requests to our service when you are on a PagePull plan, and to check how much of your monthly allowance is left.
6. Chrome Web Store Limited Use compliance
PagePull's use of information received from Google APIs and from users adheres to the Chrome Web Store User Data Policy, including the Limited Use requirements. Specifically: data is used only to provide the extraction and monitoring features you request; page content is transferred only to our own extraction service and to the Anthropic API, solely to perform an extraction you initiated, and is not retained; it is not used for advertising or any unrelated purpose; and no humans read it.
7. Data retention and deletion
Everything the extension produces — your settings, recipes, monitors and extracted tables — is stored on your device. Removing a recipe or monitor deletes it immediately, and uninstalling the extension deletes all of it, including any API key you entered.
On our side, the only records are the usage counts and page hostnames described in section 2, plus subscription details if you have a paid plan. Page content and results are never written down, so there is nothing there to delete. Usage records are kept for 12 months and then deleted. To have your usage record or subscription data erased sooner, email [email protected] and we will do it within 30 days.
8. Children
PagePull is a business productivity tool and is not directed to children under 13.
9. This website
pagepull.app is a static site hosted on Cloudflare Pages. It sets no cookies and runs no analytics or tracking scripts. Cloudflare processes standard server request logs as part of delivering and protecting the site.
10. Changes to this policy
If this policy changes, the date at the top of this page will be updated, and any material change affecting how data is handled will be described in the extension's release notes.
11. Contact
Questions, concerns, or privacy requests: [email protected].