Privacy Policy
WebstashAi — Effective May 22, 2026 · Last updated September 8, 2026
WebstashAi ("we", "our", "us") is a personal bookmark manager that lets you save web pages and text highlights, then search them later using natural language. This policy explains what data the WebstashAi browser extension and web application collect, how we use it, and what rights you have.
1. Information We Collect
a. Information You Provide
- Account information — email address, name, and profile picture provided when you sign up or log in. WebstashAi supports Google OAuth and GitHub OAuth; the specific fields received depend on the provider you choose. We also support API keys and email/password sign-in for advanced users.
- Bookmarks you save — when you click the extension icon, press a keyboard shortcut, or use the right-click menu, we capture the page URL, title, and HTML content of the current tab.
- Text highlights — when you select text and use the highlight feature, we capture the selected text, a short surrounding context, and CSS selector information to relocate the highlight on the page.
- Notes and tags — any notes or tags you manually add to saved bookmarks or highlights.
b. Information Stored Locally on Your Device
- Authentication tokens — OAuth access and refresh tokens stored in
chrome.storage.local, accessible only to the WebstashAi extension. - Offline queue — if you save a page while offline, the pending save is stored locally for retry when its owning account is signed in and a connection is available. Saves whose account cannot be established are held locally rather than assigned to the next signed-in account.
- Extension preferences — your settings and UI preferences.
- Previously saved URLs — the extension keeps a local record of URLs saved on this device to decide which pages may receive automatic saved-status checks and highlight restoration.
c. Browsing and Automatic Checks
- We do not send a general history of every page you visit, or collect keystrokes or form inputs.
- For URLs not previously saved by the extension on this device, automatic saved-status checks and highlight restoration do not send the page URL to our API.
- For a URL previously saved on this device, the signed-in extension may automatically send that URL to our API to check saved status and retrieve saved highlights. This can happen when you revisit or navigate to the page, without a new save or highlight action.
- Saving new page content or a new text highlight requires you to trigger the corresponding save or highlight action.
2. How We Use Your Information
| Purpose | Data Used |
|---|---|
| Store and organize your bookmarks | URLs, titles, page content, tags |
| Generate searchable embeddings and summaries | Page content (processed server-side by AI) |
| Provide natural-language search | Search queries matched against your saved content |
| Authenticate your account | OAuth tokens, email |
| Sync across devices | All saved data via our API |
| Maintain and improve the service | Aggregated, anonymized usage metrics |
We use your data only to provide and improve the bookmark management features described in our Chrome Web Store listing. We do not use your data for any other purpose.
3. AI Processing
WebstashAi uses AI on both Cloudflare's infrastructure (Workers AI) and Anthropic's API (Claude family). Each provider receives only the data needed for the specific feature:
- Vector embeddings for semantic search — processed by Cloudflare Workers AI from saved page content.
- Page classification, content summaries, and auto-assigned tags — processed by Anthropic (Claude) from saved page content.
- Natural-language search and chat answers — processed by Anthropic (Claude) from your search query plus relevant snippets retrieved from your own library.
- Conversational query rewriting — processed by Anthropic (Claude Haiku) from your current message plus the last few turns of conversation, so follow-ups like “what about hooks?” resolve against the right topic.
- OCR on uploaded images and PDFs — processed by Cloudflare Workers AI.
- Fallback content extraction — when a saved page blocks automated access, we may send the URL to Jina Reader to retrieve a clean text version. This is feature-gated and only fires for pages that refuse the standard fetch.
Private saved captures still use the existing Cloudflare Workers AI and Anthropic processing for summaries and embeddings. Saving a private capture does not mean no AI or no third-party processing.
Cloudflare Workers AI and Anthropic both contract with us not to retain your data for model training. The derived data (embeddings, summaries, tags, chat answers) is stored with your account and deleted when you delete the associated bookmark, conversation, or account.
When the operator has configured an observability backend, request traces — including your account user ID, internal bookmark identifiers (page IDs), your query (and any AI-rewritten form of it), the assistant's answer, URLs and titles of retrieved bookmarks, and short text snippets from those bookmarks — may be forwarded to Langfuse for debugging and quality monitoring. If you submit thumbs-up or thumbs-down feedback on an AI answer, that rating is also forwarded as a score linked to the same trace. Telemetry is processed under our legitimate interest in maintaining and improving the service; it is not used for advertising or sold to third parties. EU/EEA residents may object to this processing at any time under GDPR Article 21 by emailing [email protected] with the subject “Object to telemetry”.
The web app also sends limited first-party operational telemetry to the WebstashAi API: Core Web Vitals (LCP, INP, and CLS), a broad client-error category, a build identifier, and a normalized route template. These events do not include raw URLs, saved-page or account identifiers, error messages, stack traces, request bodies, or search prompts. They are written to Cloudflare Worker logs and, when configured, Cloudflare Analytics Engine. We do not publish an exact retention period for this operational telemetry until the active provider settings have been independently verified.
4. Data Sharing and Third Parties
Infrastructure and processors
| Provider | Purpose | Data Accessed |
|---|---|---|
| Cloudflare | API hosting, database, file storage, vector index, queue, edge cache, Workers AI (embeddings and OCR) | All saved content (encrypted in transit and at rest) |
| Anthropic | Claude API used for chat/search answers, page classification, content summaries, and conversational query rewriting | Page content snippets, your search query, conversation history, and tag/category hints |
| Jina Reader | Fallback content extraction when a saved page blocks the standard fetch (feature-gated) | The saved page URL, which Jina then fetches and converts to clean text on demand |
| ScrapeCreators | Public metadata and content enrichment for saved URLs from X (Twitter), TikTok, LinkedIn, YouTube, Reddit, Pinterest, Spotify, Apple Music, and GitHub | The saved platform URL |
| Langfuse | Observability and quality monitoring for AI features (only when the operator has configured a backend) | Account user ID, internal bookmark identifiers (page IDs), your query (and any AI-rewritten form of it), the assistant's answer, URLs and titles of retrieved bookmarks, short text snippets from those bookmarks, and thumbs-up/thumbs-down feedback ratings you submit on an answer |
| OAuth authentication for “Sign in with Google” | Email, name, profile picture (during login only) | |
| GitHub | OAuth authentication for “Sign in with GitHub” | Primary verified email and basic profile (during login only) |
| X (Twitter) | Optional X integration for users who connect an X account — pulls bookmarks/tweets you have saved on X into your WebstashAi library | X bookmarks and tweet metadata for the connected account; access tokens stored encrypted at rest |
| Resend | Transactional email delivery (verification, password reset, account notices) | Email address and message body |
When ScrapeCreators enrichment runs for a supported platform, WebstashAi sends the saved URL to ScrapeCreators for public metadata and content enrichment. We do not send browser cookies, session tokens, or raw browser HTML to ScrapeCreators. Known-restricted targets and private-collection targets skip ScrapeCreators enrichment.
Anthropic, Langfuse, and Resend each contract with us not to retain your data for model training or for purposes unrelated to providing their service to WebstashAi. Jina Reader is accessed through a key-gated public API rather than a signed data-processing agreement; we minimize our use of it to the fallback case described in Section 3, and we send it only the URL of the saved page on demand.
What we do NOT do
- We do not sell your personal data to anyone, ever.
- We do not use your data for personalized advertising.
- We do not share your data with advertising platforms, data brokers, or information resellers.
- We do not use your data to assess creditworthiness or for any form of lending decisions.
- We do not allow humans to read your saved content except: (a) with your explicit consent, (b) in aggregated/anonymized form for internal operations, (c) for security investigations, or (d) as required by law.
5. Data Security
- All data is transmitted over HTTPS/TLS.
- Data at rest is stored in Cloudflare's encrypted infrastructure.
- Authentication tokens are stored in
chrome.storage.local, which is sandboxed per-extension and inaccessible to other extensions or websites. - OAuth uses the PKCE (Proof Key for Code Exchange) flow to prevent token interception.
6. Data Retention and Deletion
- Active accounts: Your data is retained for as long as your account is active.
- AI conversations: Chats are kept for 30 days after their last message.
- Bookmark deletion: When you delete a bookmark, its content, embeddings, summaries, highlights, and notes are permanently removed from our systems.
- Account deletion: You can delete your entire account from the Settings page. This permanently removes all your data, including saved pages, highlights, notes, embeddings, and AI-generated summaries.
- Local data: Uninstalling the extension removes all locally stored data (tokens, offline queue, preferences).
7. Your Rights
All users
- Access: View all your saved data in the WebstashAi web app or extension.
- Export: Export your bookmarks and data from the Settings page.
- Delete: Delete individual bookmarks or your entire account at any time.
- Disable: Disable or uninstall the extension at any time to stop all data collection.
EU/EEA residents (GDPR)
Under the General Data Protection Regulation, you have the right to:
- Access your personal data
- Rectify inaccurate data
- Request erasure ("right to be forgotten")
- Restrict processing
- Data portability
- Object to processing
- Withdraw consent at any time
Our legal basis for processing is consent (you choose to save each page) and legitimate interest (account management, service operation). You may lodge a complaint with your local data protection supervisory authority.
Your data is processed by Cloudflare and may also be forwarded to our other processors — including Anthropic, Jina Reader, Langfuse, and Resend, plus the OAuth provider you choose to sign in with (Google or GitHub) and the optional X (Twitter) integration if you connect it. Several of these processors are US-based and their infrastructure may include servers outside the EU/EEA. Each processor maintains appropriate safeguards (such as Standard Contractual Clauses) for international data transfers.
California residents (CCPA/CPRA)
- You have the right to know what personal information we collect, use, and disclose.
- You have the right to request deletion of your personal information.
- You have the right to opt out of the sale of personal information. We do not sell your personal information.
- We will not discriminate against you for exercising any of these rights.
8. Content Script Disclosure
The WebstashAi extension includes a content script that runs on web pages you visit. This script:
- Listens for text selection events to enable the highlight capture feature
- Uses the extension's local saved-URL record to limit automatic API lookups to pages previously saved on this device
- For those pages, can request a saved-status check and retrieve existing highlights through the signed-in extension, without a new user action
- Restores saved highlight markings in the page and shows controls for highlights you choose to create; it does not inject ads
- Does not automatically upload new page content or create new saved highlights merely because you visit a page
9. Children's Privacy
WebstashAi is not directed at children under the age of 13. We do not knowingly collect personal information from children under 13. If you believe a child has provided us with personal data, please contact us and we will delete it.
10. Changes to This Policy
We may update this privacy policy from time to time. If we make material changes, we will notify you via the extension or email. The "Last updated" date at the top of this page indicates when the policy was last revised.
11. Contact Us
If you have questions about this privacy policy or want to exercise your data rights, contact us at:
- Email: [email protected]
- Website: webstashai.com