Choose the full context.
Country, subdivision, industry, product, event, place basis, person, and tax year are explicit fields. The date of the event is recorded separately.
RIGHT CONTEXT. EXPLICIT EVIDENCE.
A useful calculation belongs to a place, a time, and a situation. We’re building those facts into the rule itself, so an answer can explain where it came from.
The registry currently contains four fictional demonstration packs and zero real-world tax or regulatory packs. Existing tax tools use the rate you enter. No country’s tax, payroll, or legal rules are automatically selected.
Change a field and see whether a rule can still be applied. Every fixture below is fictional.
A subdivision needs its own matching scope.
These are independent facts. A language, currency, or industry name cannot fill them in reliably.
The review-check date is editable only to demonstrate expiry. A production approval gate must use a trusted server clock.
The event date selects the effective rule. The review-check date tests whether its evidence is still current. These are different questions.
Start included. End excluded.
REVIEWNext review due: 2027-01-01
UseCalculator synthetic-fixture test suite
Retrieved 2026-10-29 · Fictional fixture specification v1
sha256:545a1895ae60cc9875ba172d3dc16156586fd79824d3f36baeea4786084b9fb2Use a percentage you supply. This worksheet does not select, approve, or verify a tax rate for you. Its result is always labeled as your own assumption.
Open the full user-entered tax toolDot decimals · two places for amount · up to four for rate
These constraints are enforced by the prototype resolver and tested against deliberate failures. They do not certify the correctness of a real legal interpretation.
Country, subdivision, industry, product, event, place basis, person, and tax year are explicit fields. The date of the event is recorded separately.
Every rule has an effective interval and a review interval. Event dates select the applicable version; a current review gate checks whether it may be used.
Source URLs, exact source locations, retrieval dates, content fingerprints, assumptions, exclusions, and test evidence travel with each rule pack.
Missing context, overlapping versions, unreviewed drafts, retired rules, and overdue reviews block automatic resolution. No silent national or industry fallback.
A reviewed pack selects a known deterministic engine and bounded parameters. A downloaded rule cannot introduce executable JavaScript.
An exported run records its inputs, result, exact rule version, context, and review-check date. New rules do not rewrite a historical snapshot.
Real coverage needs a country-by-country and domain-by-domain release process. A source URL and a passing test are necessary pieces of evidence; specialist interpretation and maintained ownership still matter.
Include who and what the rule applies to, which place determines it, applicable dates, exemptions, thresholds, and excluded situations.
Capture the official source, its exact section, effective dates, a content fingerprint, and a responsible reviewer.
Check thresholds, rounding, periods, exemptions, and contradictory or missing inputs against approved expected answers.
Approve and publish an immutable pack. Assign a review deadline and invalidate affected coverage when source changes cannot yet be resolved.
The text below describes testing data created for this product. It is not a source of tax law, regulations, or commercial rates.
UseCalculator synthetic handling-fee fixture specification v1. Country ZZ is fictional. ZZ national retail ordinary-goods sale delivery-location business has 200 basis points. ZZ-A retail ordinary-goods sale delivery-location business has 250 basis points. ZZ-A services professional-service sale delivery-location business has 400 basis points. ZZ-B retail ordinary-goods sale delivery-location business has 300 basis points and remains draft. All tax years are not applicable. Effective interval: 2026-01-01 inclusive to 2027-01-01 exclusive. These are test fixtures, not laws, taxes, prices, or recommendations.
sha256:545a1895ae60cc9875ba172d3dc16156586fd79824d3f36baeea4786084b9fb2Synthetic fixture expected answers v1: amount 100.00 at 2.00 percent gives added amount 2.00 and total 102.00; amount 100.00 at 2.50 percent gives 2.50 and 102.50; amount 100.00 at 4.00 percent gives 4.00 and 104.00; draft ZZ-B must not resolve; half-up 0.01 at 50 percent gives 0.01. Source, scope, date, ambiguity, and review gates are independently tested.
sha256:55adf786972f1a953d93f5afb42cc102d6a746cc1efed42cc17e13f71dae939dEach SHA-256 value fingerprints the exact corresponding text shown above. A fingerprint detects content changes; it does not establish legal authority or authenticate a reviewer.
The downloadable structure uses JSON Schema 2020-12. Runtime checks additionally enforce real dates, effective intervals, exact scope matching, and review status. The rule language currently implements only a bounded percentage engine.