Data export formats

One ZIP file (APPNOTE 6.3.10; ZIP64 records only above 4 GiB or 65,535 files). Entry names are ASCII.

An archive above 50 GiB is split into volumes (…-teil-01.zip, …-teil-02.zip, …), each a complete ZIP. Volume 1 holds README.md, manifest.json, the schemas and the datasets; the other volumes hold files, and manifest.json names the volume of every file.

Data only. Since 3 October 2026 an export is started as one of two kinds, and the manifest's contents says which: all (data and files, every archive before) or data (every dataset as JSON and CSV, no files/). A data-only archive is this format without files/: the same datasets, columns, schemas and encoding; a file column still names {"path", "sha256", "bytes"} of each file as a full export holds it, manifest.json lists no files and says so in contentsNote. An archive without contents holds files.

The manifest's format is miphu-company-export and formatVersion is 1. formatRevision (1.79 since 23 October 2026) names the additions within version 1; an archive without it is 1.0. A revision is a name, not a decimal: 1.10 follows 1.9.

Revision 1.1 added the dataset crm.partners (every business partner, prospects included) and appended columns to crm.leads, crm.customers and crm.contacts; every column of 1.0 kept its place, name, type and meaning.

Revision 1.2 appended articles, stock, purchasing and personnel to team.invitations (the rights an invitation offers); every column of 1.1 kept its place, name, type and meaning.

Revision 1.3 appended task_changed and task_started to team.notification-settings (two new push switches); every column of 1.2 kept its place, name and type, and task_digest there has no effect any more (the morning summary is no longer sent). Since 28 September 2026 the stored switch is deleted and task_digest is always false; the column keeps its place, name and type.

Revision 1.4 appended legalFormName to company.profile (a legal form in the company's own wording) and access_removed_at to personnel.people (the record of a person whose access was deleted); every column of 1.3 kept its place, name, type and meaning. Since then every access has a personnel record.

Revision 1.5 added the dataset team.interface-settings (each member's table layouts in the web and the Mac app); every column of 1.4 kept its place, name, type and meaning.

Revision 1.6 added the dataset crm.erased-records (the keys of leads and contacts deleted permanently); every column of 1.5 kept its place, name, type and meaning.

Revision 1.7 added the datasets personnel.pay-agreements, personnel.pay-components and personnel.pay-requests (pay, in the owner's export only) and documents.person-releases, and appended personnel_number to personnel.people and personnel_allowed to ai.settings; every column of 1.6 kept its place, name, type and meaning.

Revision 1.8 appended first_retrieved_at to documents.person-releases (when the person first opened a released document; since then the openings under My documents are no longer logged one by one in documents.document-access), personnel_confirmed_at, personnel_confirmed_by and personnel_notice_version to ai.settings (the owner's confirmation before AI connections may read personnel data) and restrictedAt to crm.partners (a customer whose personal data were removed); every column of 1.7 kept its place, name, type and meaning.

Revision 1.9 appended ended_at to subscription.stripe-bindings (when Stripe ended the subscription; its paid time stays); every column of 1.8 kept its place, name, type and meaning.

Revision 1.10 added the manifest's contents (and contentsNote for a data-only archive) and the data-only archive; every dataset and column of 1.9 kept its place, name, type and meaning, and a full archive differs from 1.9 only by those two manifest members.

Revision 1.11 appended days_source to schedule.absences (whether a vacation's day count is the proposal from the person's working days and the public holidays, or was set by a person); every column of 1.10 kept its place, name, type and meaning, and every vacation now carries its day count.

Revision 1.12 appended cancelled_at and cancel_reason to money.expenses (a booked expense cancelled with its reason; its status is then cancelled), paymentTermDays to company.profile (the payment term in days new invoices take) and zahlungszielTage to the customers (a customer's own payment term, empty for the company's); every column of 1.11 kept its place, name, type and meaning. An expense original may now be an XML e-invoice (media_type application/xml).

Revision 1.13 appended erinnernAm to crm.tasks (when the task reminds its people by push) and task_due to team.notification-settings (the switch for that push); every column of 1.12 kept its place, name, type and meaning.

Revision 1.14 added the datasets documents.contract-terms (the terms of documents of the type contract) and documents.contract-reminders (the reminder task of each notice date), and appended document_id to crm.task-references (a task about a document); every column of 1.13 kept its place, name, type and meaning.

Revision 1.15 appended vatRegistrations to company.profile (the other countries the company is registered in for VAT, with the number where it was given); every column of 1.14 kept its place, name, type and meaning.

Revision 1.16 added the datasets money.dunning-notices (payment reminders and dunning notices with their PDF) and money.dunning-requests, and appended dunning to company.document-prefixes (the prefix of their numbers); every column of 1.15 kept its place, name, type and meaning.

Revision 1.17 added the datasets money.recurrences (recurring invoices and tasks), money.recurrence-runs (each period run, with the draft or task it created) and money.recurrence-requests; every column of 1.16 kept its place, name, type and meaning. An invoice's draft may name servicePeriod (first and last day) beside or instead of serviceDate.

Revision 1.18 added the datasets team.calendar-links (the members' calendar links, without their key) and team.calendar-setting; every column of 1.17 kept its place, name, type and meaning.

Revision 1.19 appended to documents.invoices the frozen gross amount (snapshotGross), to documents.issued-quotes the frozen currency and net amount (currency, net) and the follow-up day (follow_up_on), to documents.orders the confirmation day, title, buyer name, currency, net and gross amount as stored beside the order (confirmed_on, titel, buyer_name, currency, net, gross), and quoteFollowUpDays to company.profile; the generated values repeat what the document's own JSON holds. Every column of 1.18 kept its place, name, type and meaning.

Revision 1.20 added the dataset inventory.units (the company's own units with the standard code each stands for) and appended unit_id to inventory.articles; every column of 1.19 kept its place, name, type and meaning.

Revision 1.21 added the dataset inventory.article-units (further units of an article with their factor) and appended article_unit_id to inventory.codes, factor to inventory.order-lines and inventory.delivery-note-lines and unit_code to inventory.attribute-values; every column of 1.20 kept its place, name, type and meaning.

Revision 1.22 added the category worktime (recorded working time) with the datasets worktime.settings, worktime.confirmations, worktime.events, worktime.corrections, worktime.blocks, worktime.chains, worktime.pruned and worktime.requests, in the owner's export only; every column of 1.21 kept its place, name, type and meaning.

Revision 1.23 added the dataset money.exchange-rates (the rates kept with documents in another currency than the company's base currency); every column of 1.22 kept its place, name, type and meaning.

Revision 1.24 added the datasets schedule.holiday-days (the company's own days beside the public holidays: a day off, a half day, a closure, a holiday worked through) and schedule.holiday-requests, and appended location_country and location_region (the place of work, when it is not the company's seat) to personnel.people; every column of 1.23 kept its place, name, type and meaning.

Revision 1.25 appended sold to inventory.articles; sales_price of an article that is not sold may be empty. Every other column of 1.24 kept its place, name, type and meaning.

Revision 1.26 appended rechnungsweg to company.profile (the company's answers to the invoicing duty of its country: the threshold of a stage, and a duty lowered with its reason); every column of 1.25 kept its place, name, type and meaning. An invoice recorded from a national system carries extern (its system and when it was recorded) in its snapshot.

Revision 1.27 lets variant_key of inventory.articles hold a number part n:<value> (the value on a number axis, in that attribute's unit) beside option ids, and lets a variant have rows in inventory.bill-of-materials. Every column of 1.26 kept its place, name and type.

Revision 1.28 lets money.exchange-rates hold rates set by hand on payments (kind payment, the payment's id) and on invoice drafts; an invoice in another currency than the one its seller's country declares tax in carries steuerwaehrung (that tax total, its rate and where the rate is from) in its snapshot. Every column of 1.27 kept its place, name, type and meaning.

Revision 1.29 added the datasets inventory.workcenters (the production's work centres) and inventory.production-requests, and appended production (the right „Fertigung“) to team.memberships and team.invitations; every column of 1.28 kept its place, name, type and meaning.

Revision 1.30 added the datasets inventory.routings and inventory.routing-steps (the routing of an article: its work steps at work centres with set-up and unit times); every column of 1.29 kept its place, name, type and meaning.

Revision 1.31 added the production orders (inventory.production-settings, inventory.production-numbers, inventory.production-orders, inventory.production-order-materials, inventory.production-order-steps, inventory.production-reports), appended production_order to company.document-prefixes, and lets inventory.stock-movements hold the kinds production_consume and production_produce with the source production; every column of 1.30 kept its place, name, type and meaning.

Revision 1.32 added the dataset money.tax-exchange-rates (the rate of the tax currency set by hand on an invoice whose seller country declares tax in a currency that is neither the invoice's nor the company's base currency); every column of 1.31 kept its place, name, type and meaning.

Revision 1.33 appended production_order_id to schedule.shifts (the production order a shift is planned for); every column of 1.32 kept its place, name, type and meaning.

Revision 1.34 appended segment_from to worktime.corrections (a reference correction of one segment of a block, by the instant the segment showed as its start); every column of 1.33 kept its place, name, type and meaning.

Revision 1.35 added the datasets money.proformas (proformas of invoice drafts with their PDF) and money.proforma-requests, and appended proforma to company.document-prefixes (the prefix of their numbers); every column of 1.34 kept its place, name, type and meaning.

Revision 1.36 added the dataset documents.order-time-bookings (time booked at an order: person, day, minutes, note; a removed booking keeps its row with removed_at), and worktime.requests may hold the kind block (a block corrected in one step); every column of 1.35 kept its place, name, type and meaning.

Revision 1.37 appended responsibleId to documents.quotes (the person responsible for the quote, whose follow-ups they are); every column of 1.36 kept its place, name, type and meaning.

Revision 1.38 appended decision_reason to schedule.absences (the reason the person who decided gave with an approval or a refusal, empty when none was given); every column of 1.37 kept its place, name, type and meaning.

Revision 1.39 appended giroCode to company.profile (whether issued euro invoices with a SEPA IBAN carry a GiroCode); every column of 1.38 kept its place, name, type and meaning.

Revision 1.40 added the dataset crm.lead-sources (the company's own lead sources: key, name, archived); crm.leads quelle now holds the key of a source (website, empfehlung, google, meta, messe, telefon, sonstiges or q- and twelve hexadecimal digits, matching a row of crm.lead-sources) or is empty; every column of 1.39 kept its place, name, type and meaning.

Revision 1.41 appended scope_all to ai.connections (the connection holds all areas the person may grant, now and later), and mode there and aiAccess in team.memberships may hold full (a third level: also deleting, archiving and cancelling). It also names what audit.log holds since 12 October 2026: the appended actorKind and supportAnlass (a change SwiftLabs support made on a request of the company), whose actorId is then empty. Every other column of 1.40 kept its place, name, type and meaning.

Revision 1.42 added the quality datasets inventory.quality-defect-types, inventory.quality-plans, inventory.quality-plan-checks, inventory.inspection-numbers, inventory.inspections, inventory.inspection-checks, inventory.inspection-measurements, inventory.inspection-defects, inventory.quality-holds and inventory.quality-requests; appended quality (the quality right) to team.memberships and team.invitations and inspection (the prefix of inspection numbers) to company.document-prefixes; inventory.stock-movements may hold the source quality (a rejected quantity scrapped, held goods moved). Every column of 1.41 kept its place, name, type and meaning.

Revision 1.43 appended form, laenge_mm, breite_mm, dicke_mm, durchmesser_mm, ausschnitt_laenge_mm and ausschnitt_breite_mm to inventory.lots: the form and the measures in millimetres of a rest (material left over in production), empty for every other batch and serial number. Every column of 1.42 kept its place, name, type and meaning.

Revision 1.44 added inventory.quality-nonconformities, inventory.quality-nonconformity-team, inventory.quality-nonconformity-defects, inventory.inspection-certificates (with the issued PDF as a file), inventory.inspection-certificate-lines and inventory.inspection-certificate-inspections; appended the right to release inspection certificates to team.memberships (qualityRelease) and team.invitations (quality_release), nonconformity and certificate (their number prefixes) to company.document-prefixes, nonconformity_last and certificate_last to inventory.inspection-numbers, and nonconformity_id to inventory.quality-holds and crm.task-references. Every column of 1.43 kept its place, name, type and meaning.

Revision 1.45 added the quality datasets inventory.quality-gauges, inventory.quality-gauge-calibrations, inventory.quality-sampling-states, inventory.quality-inspection-lots, inventory.quality-control-limits, inventory.quality-supplier-evaluations and inventory.quality-settings; appended aql, inspection_level, skip_qualify and skip_frequency (the AQL sample) to inventory.quality-plans, gauge_id to inventory.quality-plan-checks, the AQL plan an inspection was opened with and its verdict (aql_value … aql_verdict) to inventory.inspections, and gauge_id and gauge_overdue to inventory.inspection-checks; sample_kind there may hold aql. Every column of 1.44 kept its place, name, type and meaning.

Revision 1.46 appended externeId to crm.partners, crm.contacts and crm.leads and external_id to inventory.suppliers, inventory.articles and personnel.people: the number of each record in the system it came from (a supplier's own, not its partner's), unique in the company without regard to case. Every column of 1.45 kept its place, name, type and meaning.

Revision 1.47 appended follow_up_task_id to documents.issued-quotes: following up a quote is a task about it (crm.tasks, crm.task-references), and follow_up_on is that task's due day while it is open. Every column of 1.46 kept its place, name, type and meaning.

Revision 1.48 appended day_rate_permitted to datev.settings: whether the tax office permits the company to convert amounts in other currencies at the day rate (§ 16 Abs. 6 Satz 3 UStG); the DATEV export reads it. Every column of 1.47 kept its place, name, type and meaning.

Revision 1.49 appended privatkunde to crm.partners: whether the customer buys as a consumer; the late-payment proposal and the national invoicing duties read it. Every column of 1.48 kept its place, name, type and meaning.

Revision 1.50 appended rate to datev.run-items: the rate a run booked an invoice in another currency at, which a later cancellation or correction of that invoice is booked at. Every column of 1.49 kept its place, name, type and meaning.

Revision 1.51 appended hourlyCostRate to company.profile and added personnel.cost-rates: the internal cost of one working hour, for the company and per person, that the post-calculation of an order prices its hours at. Every column of 1.50 kept its place, name, type and meaning.

Revision 1.52 added company.number-counters and appended period and starts to company.document-prefixes and credit_shared to company.issued-invoice-series: every number series restarts per year (or month) at the company's start value, and a record numbered from then on carries its period in number_prefix (AU-2026), its number being number_prefix, a hyphen and seq with at least four digits. Every number given before keeps its form and every column of 1.51 its place, name and type.

Revision 1.53 appended lead_day and due_day to money.recurrences: the fixed day a recurrence creates its draft or task on, and the fixed day its invoices fall due. Every column of 1.52 kept its place, name, type and meaning.

Revision 1.54 added money.invoice-payment-discounts (Skonto taken on invoices, and taken back) and appended skonto to crm.partners, skonto_units to money.invoice-payment-balances and discount_id to money.invoice-payment-requests; entry_id there is empty for a request that recorded a Skonto alone. Every column of 1.53 kept its place, name, type and meaning.

Revision 1.55 added recurring expenses: money.recurrences takes kind expense and appends source_expense_id, supplier_id and confirm, money.recurrence-runs appends expense_id. Every column of 1.54 kept its place, name and type; kind, due_day and template gained the expense case beside their former meaning.

Revision 1.56 added optional and alternative quote positions: a position in a quote's draft and frozen snapshot may carry art (optional or alternativ) and alternativZu (the id of the ordinary position it replaces); neither counts in the quote's calculation. The decision of documents.issued-quotes may carry auswahl, the positions that apply. No column changed.

Revision 1.57 added prorated first periods and price changes of recurring invoices: money.recurrences appends prorate and service_start, and the dataset money.recurrence-price-changes holds the lines in force from a day on. Every column of 1.56 kept its place, name and type; anchor of a prorated rule is the first day of the calendar unit its service begins in.

Revision 1.58 added the Sammelrechnung and cumulative advances: an invoice's draft and frozen snapshot may carry sammel (the orders and delivery notes it bills, each with number, date and its line ids), abrechnung.auftraege (every order a Schlussrechnung settles) and abrechnung.kumuliert (a cumulative Abschlag: the work completed to date and the earlier advances it is billed less of); a deduction may carry kumuliert. A datev.run-items row may have kind anzahlung, a payment on an advance whose tax went over with it. No column changed.

Revision 1.59 added offsets and refunds of a credit: money.invoice-payments takes kind refund (negative, its correction positive), the dataset money.invoice-payment-allocations holds offsets with another invoice of the same customer, money.invoice-payment-balances appends allocated_units and money.invoice-payment-requests allocation_pair. Every column of 1.58 kept its place, name and type.

Revision 1.60 added the dataset money.dunning-stops (an invoice or a customer held out of dunning, with reason and last day). The fee, interest and flat-rate compensation of an issued money.dunning-notices row now count to the invoice's open item and receivable; no column changed.

Revision 1.61 added the discount on the whole document: the draft of a quote, an order or an invoice (in documents.invoices and the sales datasets, and in a frozen snapshot) may carry belegRabatt (art prozent or betrag, wert, grund); its calculation then names totals.lineTotal (the lines' net), totals.allowance (kind, value, reason and the amount taken off) and each tax group's allowance - the group's basis is already less it, so totals.net is the lines' net less the discount. No column changed.

Revision 1.62 added the retention of VOB/B § 17 and the Teilschlussrechnung: the datasets documents.order-retention-terms (an order's default retention) and money.invoice-retention-releases (a retention released by a guarantee, newest revision holds). An invoice's draft and frozen snapshot may carry abrechnung.teil (a Teilschlussrechnung), abrechnung.abschlaege (the advances it deducts), abrechnung.teilschluesse (the parts a Schlussrechnung follows) and abrechnung.einbehalt, its calculation totals.retention (held back of the amount due). Every column of 1.61 kept its place, name and type.

Revision 1.63 added SEPA direct debit: the datasets money.sepa-creditor (the creditor identifier), money.sepa-mandates (the customers' mandates), money.sepa-runs (a run and the pain.008 file's message id) and money.sepa-collections (each invoice a run collects, with mandate, sequence and amount). An invoice draft (in documents.invoices) may carry lastschrift (mandatId, referenz, glaeubigerId, iban, art). No column changed.

Revision 1.64 added the bank accounts: the datasets money.payment-accounts (bank accounts and cash boxes) and money.payment-account-requests; account_id appended to money.invoice-payments and money.expense-payments (the account a payment went through). An expense's fields (in money.expenses) may carry taxCase, selfAssessedRate, serviceDate and servicePeriod; an invoice line in a draft or snapshot leistungsart (lieferung, leistung) and vorsteuerabzug. Every column of 1.63 kept its place, name and type.

Revision 1.65 added the VAT settings: the datasets money.tax-settings (every saved version of soll or ist, the filing period, the deadline extension and the special prepayment) and money.tax-settings-requests. An expense's fields may carry supplyType (goods, services) and inputTaxDeduction. No column changed.

Revision 1.66 added the filed VAT returns: the datasets money.tax-filings (each filing with its form values and the one it corrects), money.tax-filing-items (the documents behind it) and money.tax-check-marks. No column changed.

Revision 1.67 added online acceptance of quotes: the datasets documents.quote-links (a quote's link for its customer, with the PDF frozen with it under files/documents/quote-links/ and its SHA-256, never the link's key), documents.quote-link-decisions (the customer's decision: outcome, name, reason, positions chosen, total, network and device class) and documents.quote-link-requests. audit.log gained customerLink and the actor kind customer, documents.quote-requests gained actorKind, quoteLinkId and memberId; a quote's decision (in documents.issued-quotes) may carry quelle, name and linkId. No column changed.

Revision 1.68 added online payment through Stripe: the datasets money.online-payment-connections (the connection, new or the company's own Stripe account, its status as Stripe reported it, without Stripe's account id), money.online-payment-settings and money.online-payment-requests; anbieter appended to money.payment-accounts, whose kind may be zahlungsdienst (the Stripe balance the connection keeps). Every column of 1.67 kept its place, name and type.

Revision 1.69 appended bezuege to money.tax-check-marks: the records (ausgabe|<id>, rechnung|<id>) a check referred to when it was marked as not applying („trifft nicht zu“), empty for a mark taken back and for marks recorded before. No column changed.

Revision 1.70 appended quoteLink to company.profile (whether issuing a quote makes its link for online acceptance and prints it with its QR code).

Revision 1.71 added approvals with limits: the datasets documents.approval-rules, documents.approval-requests, documents.approval-decisions, documents.approval-commands and team.approval-deputies. No column changed.

Revision 1.72 added the withdrawal of a consumer's online acceptance (§ 356a BGB): the datasets company.withdrawal-terms (the company's Widerrufsbelehrung and whether consumers may ask for an early start) and documents.quote-link-withdrawals (who withdrew, when, the address of the acknowledgement and how its mail went). documents.quote-links gained withdrawal, withdrawal_text, withdrawal_text_sha256 and early_start (what a consumer's link froze), documents.quote-link-decisions gained early_start. No column changed.

Revision 1.73 added an invoice's payment link: the datasets money.online-payment-links (the link, its key version, who made it, how often it was opened, how it ended; without the token's hash) and money.online-payment-checkouts (the Stripe payment pages a link opened: session id, amount in minor units, currency, cash discount stage, status; without their address). No column changed.

Revision 1.74 added the bookings from Stripe: the datasets money.online-payments (each payment through a payment link: method, amount, status, when Stripe credited it, Stripe's fee), money.online-payment-refunds, money.online-payment-disputes and money.online-payment-payouts. money.invoice-payments gained source (manual or online), provider_ref (the Stripe object) and cause (on an online refund: dispute or refund), money.invoice-payment-discounts gained source and provider_ref; actor_id is empty on an online row. A dispute's withdrawal is a refund with cause dispute. actorKind of audit.log may be provider (a change booked from Stripe; actorId is then empty). Every column of 1.73 kept its place, name and type.

Revision 1.75 added the company protocol: the datasets audit.protocol-events (both hash chains, each event with its hash and its predecessor's, so the chain can be recomputed outside miphu; without the network of a sign-in), audit.protocol-chains (the heads) and audit.protocol-settings (the retention). No column changed.

Revision 1.76 added the books: the datasets money.ledger-entries and money.ledger-lines (the journal, each entry with its link of the company's hash chain and its predecessor's hash, so the chain can be recomputed outside miphu), money.ledger-chain, money.ledger-settings, money.ledger-accounts, money.ledger-account-map, money.ledger-partner-accounts (every version kept), money.ledger-fiscal-years, money.ledger-fiscal-year-events, money.ledger-periods, money.ledger-period-events, money.ledger-runs and money.ledger-requests. module of company.modules may be buchfuehrung (the books). No column changed.

Revision 1.77 added the payment accounts in the books: money.ledger-payment-accounts (the general ledger account each payment account was given) and money.ledger-balance-confirmations (balances confirmed with their statement). No column changed.

Revision 1.78 added travel expenses: the datasets money.expense-claims (the expenses a person paid and is reimbursed for), money.trips (business trips with what was entered and the per diems and mileage calculated from it, frozen at submission), money.trip-requests and company.travel-rates (the company's own rates). starts_at and ends_at of money.trips are local times (see Encoding). No column changed.

Revision 1.79 lets rechnungsformat of crm.customers and crm.partners hold xrechnung (a public body without a Leitweg-ID: every invoice to it is an XRechnung), and automatisch now also means an e-invoice where the company's seat makes the hybrid e-invoice the standard. No column changed.

An archive above 50 GiB arrives in volumes of at most 50 GiB, each a complete ZIP and its own download; a 2 TB export is 38 volumes. The export page lists every volume with its size. In the browser each volume is an ordinary download straight to the downloads folder: an interrupted download continues from the browser's download list or starts again from the export page. The apps load in pieces and continue after an interruption.

Exports are built in the background in two lanes. Before it starts, an export's size is estimated from the database and the sizes of its files: up to 5 GiB it runs in the small lane (up to eight at a time across the service), above it in the large lane (up to two at a time); a large export never takes the place of a small one. No export starts while the exports being built or ready for download would take up more than 20 TiB of storage, and large ones leave 1 TiB of it to the small lane. Until it starts an export shows „Wartet auf den Start“ („Waiting to start“) and why it waits; it starts by itself, and a mail says when it is ready. A large export takes hours to build and stays downloadable for seven days.

A company starts one export at a time: an export with files once in 24 hours, a data-only export up to five times in 24 hours. A failed export does not count, and the export page says when the next one is possible.

• Text: UTF-8 everywhere.
• JSON: keys are the column names below, in this order. Times are UTC in ISO 8601 with a Z (2026-09-15T08:30:00.123Z); dates are YYYY-MM-DD; amounts (numeric) and 64-bit integers (bigint) are decimal strings, so no value loses precision; JSON columns are embedded as stored.
• CSV: RFC 4180, comma, CRLF, a header row with the column names, every value in double quotes, no byte order mark. An empty field is null. JSON columns and lists appear as JSON text. JSON is authoritative where CSV cannot tell an empty text from a missing value. Values are written unchanged; a spreadsheet should import them as text.
• Files: a column that holds a file is written as {"path", "sha256", "bytes"} in JSON and as the path in CSV; the file is stored under that path.
• Integrity: sha256 in the manifest and in file references is the SHA-256 of the uncompressed bytes, as lower-case hex.
• Times (timestamp … with time zone) are instants, written in UTC. The SQL type says the precision: timestamp(3) with time zone has three fractional digits (milliseconds, …:00.123Z), timestamp with time zone six (microseconds, …:00.123456Z). Dates (date) are calendar days of the company's time zone (company.profile.timezone), with no time and no zone. Local times (timestamp … without time zone) are a date and a time of day as they were entered, with no zone and never converted: 2026-10-23T08:30:00.000000.
• Amounts come in two forms, and the column says which. A numeric column named units or net_units is an integer number of the currency's minor units: the amount is units / 10^minor, where minor is the currency's ISO 4217 exponent (2 for EUR and USD, 0 for JPY) — in the same dataset where it has a minor column, otherwise the exponent of the currency the description names; 4001 in EUR is 40.01 EUR. Every other amount (prices, totals, amount, value, the amounts inside JSON documents) is a decimal in major units of its currency (40.01).
• People. A column that names a person (actor_id, created_by, owner…, assignee…, …By, account_id) holds the account's id (team.accounts.id). A column named membership_id holds the membership's id (team.memberships.id); team.memberships.accountId joins the two. An account belongs to one company, so both name the same person.
• Deleted, archived and ended records are included, and a column says so: deletedAt (customers, contacts, leads, notes, tasks), archivedAt and archived_at (archived), trashed_at (a document in the trash), purged_at (a file whose content is gone), removed_at (a planned cash item taken out of the plan), revoked_at (AI connections, share links, product feeds, complimentary access), released_at and releasedAt (a released stock reservation; a membership whose company was deleted), cancelled_at (shifts, stock counts), disabledAt, anonymisedAt and anonymised_at (accounts and personnel records), and for records that end by their state a status of cancelled.

Version 1 is frozen since 2026-09-26. No column or dataset of version 1 is renamed, removed or changes its type or its meaning; its descriptions may grow, and datasets and columns may be added. Anything else starts version 2, and this register keeps the description of version 1. Some names and values stay as they were first published even where the product now says it differently (German column names and German status values in crm.* and documents.quotes, companyId beside company_id); where the AI connection (MCP) names a value differently, the column's description gives its code.

The datasets personnel.pay-agreements, personnel.pay-components, personnel.pay-requests, subscription.source, subscription.seat-control, subscription.apple-status, subscription.apple-bindings, subscription.apple-transactions, subscription.apple-renewals, subscription.apple-purchase-intents, subscription.operator-grants, subscription.stripe-bindings, subscription.complimentary-access, subscription.apple-sandbox, subscription.cloud-status, subscription.cloud-bindings, subscription.cloud-purchase-intents, subscription.stripe-cloud, subscription.stripe-changes, documents.email-addresses, crm.objection-salt, worktime.settings, worktime.confirmations, worktime.events, worktime.corrections, worktime.blocks, worktime.chains, worktime.pruned, worktime.requests, personnel.cost-rates are in the owner's export only; an administrator's export lists them in manifest.json under excluded.ownerOnlyDatasets.

company.profile

The company: name, address, tax and invoice details, currency, time zone, trial and seat facts.

Files: data/company/profile.json, data/company/profile.csv, schema/company/profile.schema.json

company.legal-acceptances

Acceptances of the terms of use and the data processing agreement: which account accepted which version, and when.

Files: data/company/legal-acceptances.json, data/company/legal-acceptances.csv, schema/company/legal-acceptances.schema.json

company.invoice-number-series

Counter of draft invoice numbers.

Files: data/company/invoice-number-series.json, data/company/invoice-number-series.csv, schema/company/invoice-number-series.schema.json

company.issued-invoice-series

Number series of issued invoices: prefix, next number, whether it is locked; since RL2 its settings — the prefix of a series of its own for corrections (credit_prefix) and yearly series (yearly). Since 1.52 the counters are in company.number-counters; next_number and yearly hold their last state before that.

Files: data/company/issued-invoice-series.json, data/company/issued-invoice-series.csv, schema/company/issued-invoice-series.schema.json

company.quote-number-series

Number series of issued quotes per year until 1.52; since then quotes count in company.number-counters.

Files: data/company/quote-number-series.json, data/company/quote-number-series.csv, schema/company/quote-number-series.schema.json

company.quote-draft-counter

Counter of quote drafts.

Files: data/company/quote-draft-counter.json, data/company/quote-draft-counter.csv, schema/company/quote-draft-counter.schema.json

company.tags

The company's tags (Schlagworte) with colour; archived tags are included and flagged.

Files: data/company/tags.json, data/company/tags.csv, schema/company/tags.schema.json

company.tag-assignments

Which tag is set on which lead, customer or contact.

Files: data/company/tag-assignments.json, data/company/tag-assignments.csv, schema/company/tag-assignments.schema.json

company.tag-requests

Requests that created, changed or assigned tags.

Files: data/company/tag-requests.json, data/company/tag-requests.csv, schema/company/tag-requests.schema.json

company.creation-request

The request that created the company.

Files: data/company/creation-request.json, data/company/creation-request.csv, schema/company/creation-request.schema.json

company.demo-registration

Marks a demonstration company. Empty for every real company.

Files: data/company/demo-registration.json, data/company/demo-registration.csv, schema/company/demo-registration.schema.json

team.accounts

Name, e-mail address and state of every account that has or had access to the company.

Files: data/team/accounts.json, data/team/accounts.csv, schema/team/accounts.schema.json

team.memberships

Access of each account: role, CRM, finance and planning rights, title, rank, status.

Files: data/team/memberships.json, data/team/memberships.csv, schema/team/memberships.schema.json

team.departments

Departments.

Files: data/team/departments.json, data/team/departments.csv, schema/team/departments.schema.json

team.department-assignments

Which member belongs to which department.

Files: data/team/department-assignments.json, data/team/department-assignments.csv, schema/team/department-assignments.schema.json

team.department-requests

Requests that changed departments.

Files: data/team/department-requests.json, data/team/department-requests.csv, schema/team/department-requests.schema.json

team.avatars

Profile pictures of members, as JPEG files with a thumbnail.

Files: data/team/avatars.json, data/team/avatars.csv, schema/team/avatars.schema.json

team.avatar-requests

Requests that set or removed profile pictures.

Files: data/team/avatar-requests.json, data/team/avatar-requests.csv, schema/team/avatar-requests.schema.json

team.access-requests

Requests that granted or changed access.

Files: data/team/access-requests.json, data/team/access-requests.csv, schema/team/access-requests.schema.json

team.notification-settings

Push notification switches of each member: which task events reach their devices. No row means every switch is on.

Files: data/team/notification-settings.json, data/team/notification-settings.csv, schema/team/notification-settings.schema.json

crm.customers

Customers. Deleted customers are included and flagged by deletedAt.

Files: data/crm/customers.json, data/crm/customers.csv, schema/crm/customers.schema.json

crm.contacts

Contacts, with their customer.

Files: data/crm/contacts.json, data/crm/contacts.csv, schema/crm/contacts.schema.json

crm.leads

Leads in every stage, with lost reason and company details.

Files: data/crm/leads.json, data/crm/leads.csv, schema/crm/leads.schema.json

crm.notes

Notes on leads, customers and contacts. Archived notes are included and flagged by archivedAt.

Files: data/crm/notes.json, data/crm/notes.csv, schema/crm/notes.schema.json

crm.tasks

Tasks with assignee, due date, status and the customer, contact, lead or department they refer to.

Files: data/crm/tasks.json, data/crm/tasks.csv, schema/crm/tasks.schema.json

crm.task-references

Quotes, orders, invoices, expenses and documents a task refers to.

Files: data/crm/task-references.json, data/crm/task-references.csv, schema/crm/task-references.schema.json

crm.lead-requests

Requests that changed leads.

Files: data/crm/lead-requests.json, data/crm/lead-requests.csv, schema/crm/lead-requests.schema.json

crm.note-requests

Requests that changed notes.

Files: data/crm/note-requests.json, data/crm/note-requests.csv, schema/crm/note-requests.schema.json

crm.record-requests

Requests that changed customers, contacts and tasks.

Files: data/crm/record-requests.json, data/crm/record-requests.csv, schema/crm/record-requests.schema.json

crm.import-batches

Imports from an earlier system: source snapshot, mapping and reconciliation.

Files: data/crm/import-batches.json, data/crm/import-batches.csv, schema/crm/import-batches.schema.json

crm.import-records

Each imported record with its original values and what became of it.

Files: data/crm/import-records.json, data/crm/import-records.csv, schema/crm/import-records.schema.json

crm.import-archive

Imported history entries the company archived.

Files: data/crm/import-archive.json, data/crm/import-archive.csv, schema/crm/import-archive.schema.json

crm.history-batches

Imports of business history.

Files: data/crm/history-batches.json, data/crm/history-batches.csv, schema/crm/history-batches.schema.json

crm.history-records

Imported business history entries (activities of the earlier system).

Files: data/crm/history-records.json, data/crm/history-records.csv, schema/crm/history-records.schema.json

documents.branding

The company's brand colour and the colours printed from it, and which logo is current.

Files: data/documents/branding.json, data/documents/branding.csv, schema/documents/branding.schema.json

documents.branding-logos

Company logos as PNG files: the current one and every earlier one an issued document still prints.

Files: data/documents/branding-logos.json, data/documents/branding-logos.csv, schema/documents/branding-logos.schema.json

documents.branding-requests

Requests that changed the logo or the brand colour.

Files: data/documents/branding-requests.json, data/documents/branding-requests.csv, schema/documents/branding-requests.schema.json

documents.presentation-records

What each issued quote, order and invoice was printed with, frozen at issue: branding, layout, issuer details, the customer's state and document language.

Files: data/documents/presentation-records.json, data/documents/presentation-records.csv, schema/documents/presentation-records.schema.json

documents.quotes

Quotes with draft, calculation, status and archive date.

Files: data/documents/quotes.json, data/documents/quotes.csv, schema/documents/quotes.schema.json

documents.issued-quotes

Issued quotes: number, frozen content, decision, the PDF as a file.

Files: data/documents/issued-quotes.json, data/documents/issued-quotes.csv, schema/documents/issued-quotes.schema.json

documents.quote-origins

Where a quote was copied from.

Files: data/documents/quote-origins.json, data/documents/quote-origins.csv, schema/documents/quote-origins.schema.json

documents.quote-requests

Requests that changed quotes.

Files: data/documents/quote-requests.json, data/documents/quote-requests.csv, schema/documents/quote-requests.schema.json

documents.orders

Orders with draft, calculation, frozen content and the PDF as a file.

Files: data/documents/orders.json, data/documents/orders.csv, schema/documents/orders.schema.json

documents.order-requests

Requests that changed orders.

Files: data/documents/order-requests.json, data/documents/order-requests.csv, schema/documents/order-requests.schema.json

documents.invoices

Invoices and cancellation invoices: draft, frozen content with its hash, number, status.

Files: data/documents/invoices.json, data/documents/invoices.csv, schema/documents/invoices.schema.json

documents.invoice-files

The issued PDF and, for e-invoices, the XML of each invoice as files, with their hashes.

Files: data/documents/invoice-files.json, data/documents/invoice-files.csv, schema/documents/invoice-files.schema.json

documents.einvoice-validation

Validation receipt of each e-invoice: hashes of PDF and XML, profiles and tool versions.

Files: data/documents/einvoice-validation.json, data/documents/einvoice-validation.csv, schema/documents/einvoice-validation.schema.json

documents.invoice-requests

Requests that created, issued or cancelled invoices.

Files: data/documents/invoice-requests.json, data/documents/invoice-requests.csv, schema/documents/invoice-requests.schema.json

documents.imported-invoice-batches

Imports of invoices from an earlier system.

Files: data/documents/imported-invoice-batches.json, data/documents/imported-invoice-batches.csv, schema/documents/imported-invoice-batches.schema.json

documents.imported-invoices

Invoices imported from an earlier system.

Files: data/documents/imported-invoices.json, data/documents/imported-invoices.csv, schema/documents/imported-invoices.schema.json

documents.imported-invoice-lines

Lines of imported invoices.

Files: data/documents/imported-invoice-lines.json, data/documents/imported-invoice-lines.csv, schema/documents/imported-invoice-lines.schema.json

documents.imported-invoice-files

Documents attached to imported invoices, as files.

Files: data/documents/imported-invoice-files.json, data/documents/imported-invoice-files.csv, schema/documents/imported-invoice-files.schema.json

documents.imported-document-batches

Imports of documents from an earlier system.

Files: data/documents/imported-document-batches.json, data/documents/imported-document-batches.csv, schema/documents/imported-document-batches.schema.json

documents.imported-documents

Imported documents and what they belong to.

Files: data/documents/imported-documents.json, data/documents/imported-documents.csv, schema/documents/imported-documents.schema.json

documents.imported-document-files

The imported document files.

Files: data/documents/imported-document-files.json, data/documents/imported-document-files.csv, schema/documents/imported-document-files.schema.json

documents.imported-quote-parents

Quotes of the earlier system that imported documents belong to.

Files: data/documents/imported-quote-parents.json, data/documents/imported-quote-parents.csv, schema/documents/imported-quote-parents.schema.json

money.invoice-payments

Recorded payments on invoices, including corrections, and since 1.59 refunds of a credit paid back to the customer.

Files: data/money/invoice-payments.json, data/money/invoice-payments.csv, schema/money/invoice-payments.schema.json

money.invoice-payment-balances

Paid amount per invoice.

Files: data/money/invoice-payment-balances.json, data/money/invoice-payment-balances.csv, schema/money/invoice-payment-balances.schema.json

money.invoice-payment-requests

Requests that recorded payments.

Files: data/money/invoice-payment-requests.json, data/money/invoice-payment-requests.csv, schema/money/invoice-payment-requests.schema.json

money.expenses

Expenses with supplier, amounts, tax and category; archived expenses are included and flagged, as are their review status (recorded/reviewed/booked) and its timestamps.

Files: data/money/expenses.json, data/money/expenses.csv, schema/money/expenses.schema.json

money.expense-requests

Requests that changed expenses.

Files: data/money/expense-requests.json, data/money/expense-requests.csv, schema/money/expense-requests.schema.json

money.expense-originals

Uploaded receipts (PDF, PNG, JPEG) as files, including replaced ones.

Files: data/money/expense-originals.json, data/money/expense-originals.csv, schema/money/expense-originals.schema.json

money.expense-original-state

The receipt file currently attached to each expense.

Files: data/money/expense-original-state.json, data/money/expense-original-state.csv, schema/money/expense-original-state.schema.json

money.expense-original-requests

Requests that attached or removed receipts.

Files: data/money/expense-original-requests.json, data/money/expense-original-requests.csv, schema/money/expense-original-requests.schema.json

money.expense-payments

Recorded payouts of expenses, including corrections.

Files: data/money/expense-payments.json, data/money/expense-payments.csv, schema/money/expense-payments.schema.json

money.expense-payment-balances

Paid amount per expense.

Files: data/money/expense-payment-balances.json, data/money/expense-payment-balances.csv, schema/money/expense-payment-balances.schema.json

money.expense-payment-requests

Requests that recorded payouts.

Files: data/money/expense-payment-requests.json, data/money/expense-payment-requests.csv, schema/money/expense-payment-requests.schema.json

money.cash-openings

Opening cash balance per currency for the cash plan.

Files: data/money/cash-openings.json, data/money/cash-openings.csv, schema/money/cash-openings.schema.json

money.cash-plan

Planned incoming and outgoing amounts (Planposten).

Files: data/money/cash-plan.json, data/money/cash-plan.csv, schema/money/cash-plan.schema.json