Independent U.S. tariff referenceNot affiliated with the U.S. government

Data handling

Keep business records under your control.

Application behavior for this release. Hosting-provider retention is separate.

Files and calculation inputs

The on-page refund workspace parses selected files and calculates in your browser. It does not submit a claim, connect to ACE, or save your file in an account.

Separate upload, calculation, and export API endpoints process the data sent to them on the server. This application does not intentionally persist those request bodies, but hosting infrastructure may retain request metadata. Do not assume every use of the service is browser-only.

Downloads remain on your device and may contain sensitive business information. Review them before sharing.

Research links

Search and comparison use URL parameters. HTS codes, origins, and dates in a shared link are visible to its recipient and may appear in browser history or server request logs.

Do not put importer numbers, account details, or confidential descriptions into research links or public issue reports.

Duty estimator

The duty worksheet sends the HTS code, origin and date to the lookup API. Customs values and value comparisons stay in browser memory; they are not added to the request or shared link. Reloading clears those values. Worksheet downloads include the amounts you entered.

Optional usage measurement

Product analytics is disabled unless explicitly enabled in the deployment configuration. When enabled, the application permits limited workflow properties such as tool type, row counts, and error counts. Financial amounts, entry numbers, uploaded rows, and query text are excluded from those event properties.

Automatic page capture, session recording, and automatic exception capture are disabled. Client analytics uses in-memory identification rather than persistent browser storage. Server completion events are operational counts, not counts of unique people or businesses.

This page does not promise that network providers receive no IP address or request metadata. Provider configuration and retention need separate review before production deployment.

Public usage counts

The site keeps simple counts of how often it is used and publishes them on the usage counts page. When a page opens or a tool finishes, the page sends the name of that event, one of: page view, visit, tariff search, entry comparison, duty worksheet, refund estimate, entry file review, workpaper download. The site adds one to that day’s count for the name.

The counting request carries no HTS code, origin, date, customs value or file content. It is sent without cookies and without a referrer, so it does not include the address of the page you are on. The counter keeps no IP address, cookie or identifier, and nothing is stored in your browser for it. A visit is recognized from the referring address your browser already reports; that check happens in the page and the address is not sent. Like every request, the counting request reaches the hosting provider, which may keep request metadata as described on this page. Automated browsers that identify themselves are not counted.

Product watchlists

Saving a product stores its HTS code, origin, optional label, topic selections, and notice read state in local storage on this browser. The watchlist is not uploaded to an account or included in analytics events. Other users of this browser profile can access it. Clearing site data removes it; private-browsing settings may prevent persistence.

Export and import create or read a local JSON backup. Review that file before sharing it. Opening a historical search sends its code, origin, and date as URL parameters as described above. Watchlists do not provide cross-device synchronization, background checks, or email alerts.

Manage or clear local watchlist →

External services

The site requests web fonts from Google Fonts. Following an official-source or GitHub link takes you to another service with its own policies. This release does not offer accounts or automatic refund tracking.

Coverage and corrections →

Section 232 entry workpapers

The qualification panel holds answers, client references, preparer details and document references in browser memory. It does not upload them, add them to search URLs, save them to browser storage or send them to analytics. A reload clears the working record; the lookup code, origin and date still use the search URL described above.

HTML, CSV and JSON downloads include the private information you enter. No supporting documents are uploaded or authenticated. A preparer name is a user-supplied label, not a verified identity or professional approval. The separate research-snapshot download does not include these qualification answers.

Page last updated . Tariff schedule in the tools: 2026 HTS Revision 23, checked November 10, 2026. Coverage & methodology