A single tool, for a single refund window.
un-tariff was built for one purpose: to help US importers recover the IEEPA duties they paid under tariffs the Supreme Court invalidated on February 20, 2026. That's it. No platform ambitions. No add-on modules. One product, one refund channel, one time-boxed window.
Why this exists
When SCOTUS ruled in Learning Resources, Inc. v. Trump, CBP stood up the CAPE refund channel as the exclusive path for importers to get their IEEPA duties back. CAPE is administratively straightforward — a CSV of entry numbers — but mechanically unforgiving: wrong HTS classification, wrong interest walk, wrong CSV encoding, and the filing gets rejected at a time (close to the 80-day cliff) when rejections hurt.
Customs brokers can do this work, but they typically charge a percentage of the recovered refund — substantial on a six- or seven-figure claim. Trade law firms bill partner hours against a filing that, for most importers, does not need a partner. And DIY is a spreadsheet project that runs on everyone's patience and produces an audit trail no one wants to defend.
un-tariff is the narrow middle: software that runs every CBP CAPE rule automatically, generates the validated CSV, traces every computation to a rule ID and a reference-data version, and gets out of the way. You upload to the ACE Portal yourself.
What we believe
Traceability over brevity
Every classification, every interest computation, every validation outcome logs which rule fired, which reference-data version it consulted, and which source row it came from. When CBP asks for reconciliation six months after filing, the answer is already documented.
Local validation before external submission
CBP's VAL-F / VAL-E / VAL-I rules run locally before the CSV is downloadable. A file that would fail at upload does not get produced. Rejections at the ACE Portal should be environmental (ACH payer ID, filer code) not content.
Reference data is versioned, not overwritten
The IEEPA HTS registry, the CBP quarterly interest rates, and the CAPE excluded-entry types are all stored with effective-from / effective-to dates. Historical classifications remain reproducible even as CBP updates guidance.
Generation is separate from submission
un-tariff produces files. Humans upload them. The separation is intentional: the software prepares, the importer decides.
Scope
Phase 1 handles CAPE-eligible entries within the 80-day cliff. Out of scope:
- Protest filings (19 CFR 174) for entries past the cliff
- Anti-dumping / countervailing duty interactions
- Classification disputes that require argument
- USMCA exclusion reconciliation (flagged, not handled)
For those, engage counsel or a specialist broker. un-tariff vs trade law firm walks through where the boundary falls.
The end state
un-tariff is a sunset product. When CBP closes CAPE (projected late 2026 / early 2027 depending on Phase 2 rollout), the tool retires. We're not building a long-term business here — we're building for a specific administrative window. That shapes every product decision: we optimize for getting the filing right the first time, not for retention or expansion.
Who we are
un-tariff is built by a small team with backgrounds in customs compliance software, tax and duty reconciliation systems, and trade-data infrastructure. Our primary customer (and internal dogfooding target) is a mid-size US textile importer — which is why the edge cases for textile / apparel / knit-fabric entries are particularly well modeled.
We are not customs brokers. We are not attorneys. We do not file on anyone's behalf. Our job is to make sure the CSV that gets uploaded is correct.
Contact
General questions: hello@un-tariff.com. Support for active filings: support@un-tariff.com. Security disclosures: see our security.txt.