The company registry MCP: brreg orgnr, Companies House, Bolagsverket organisationsnummer.
Scored across 7 tools
The core tools are clearly distinct, but `fetch` duplicates `lookup_company` plus `company_deadlines`, and `search` duplicates `search_company`; these deliberate aliases create overlap. The descriptions explain when to use which, so an agent can disambiguate, but the boundary is not as clean as it could be.
Most core tools follow a verb_noun pattern (`lookup_company`, `search_company`, `validate_company_id`, `list_countries`), but `company_deadlines` is a noun phrase and `fetch`/`search` are bare verbs. The convention is readable but not consistently applied.
Seven tools is a reasonable size for a registry domain, and the five core operations each earn their place. However, two of the seven are redundant connector aliases, so the effective tool count is smaller than advertised.
The domain is read-only company-registry data, and the surface covers the full workflow: discover supported countries, validate identifiers, search by name, fetch full company reports, and compute filing deadlines. Missing capabilities like Sweden name search are upstream register limitations rather than gaps in the tool set.
Give the next occurrence of each statutory filing deadline a company faces.
country="NO" covers the Norwegian calendar (Regnskapsregisteret, Skatteetaten) for a
company looked up by organisasjonsnummer (orgnr, org.nr) in Brønnøysundregistrene /
Enhetsregisteret (brreg): årsregnskap, generalforsamling, skattemelding,
aksjonærregisteroppgaven, mva-melding, a-melding. country="GB" covers the two
Companies House obligations for a company number (CRN): the annual accounts filing and
the confirmation statement (CS01). country="SE" covers the two Swedish obligations of
an aktiebolag (AB) or ekonomisk förening (EK) looked up by organisationsnummer at
Bolagsverket: the ordinary general meeting (ordinarie bolagsstämma / årsstämma) at six
months from the financial year end, and the annual report (årsredovisning) at seven,
where the late-filing fee (förseningsavgift) begins.
Pass today (YYYY-MM-DD) for a reproducible answer; it defaults to the server's
current UTC date. Quote due_date, not statutory_date, and quote each deadline's
applies_because rather than presenting a date as unconditional fact — that sentence
carries the legal form or flag the date rests on, its statute, any assumption still in
it, and for the UK whether it is Companies House's own figure or one computed here.
days_until goes negative for a filing Companies House still shows as overdue. Swedish
dates assume a financial year ending 31 December unless you pass include=["filings"],
which substitutes the year end of the last filed annual report where Bolagsverket's
document list holds one; the filing date is an outer limit regardless, since a company
whose general meeting was earlier must file earlier. An empty deadlines list is a real
answer — a bankrupt, deleted or compulsorily-liquidated entity, a branch/sub-unit, or
any company whose status is not active — and notes explains why.
registry://rules/{country} carries each country's full deadline rules, roll-forward
treatment and legal sources. rules_last_reviewed names the date this country's
statutes and day-count arithmetic were last checked against the law — a deadline
computed long after that date should be re-verified before anyone acts on it.
Errors are the {"error": {"code", "message", "hint"}} envelope this server's
instructions set out code by code; hint names the next call. This tool looks the
entity up first, so any lookup_company error code can surface here too.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape. | |
| today | No | Date to compute deadlines from, YYYY-MM-DD; defaults to the server's current UTC date. Anything else is a bad_request naming the format. | |
| country | No | ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set. | NO |
| include | No | Attachment names that can change a *computed* deadline — narrower than lookup_company's include. Today only 'filings': one extra upstream request for the entity's filing history, supplying a real financial year end where 31 December would otherwise be assumed. Empty by default; Norway and the United Kingdom accept it and it changes nothing for them today. Any other value — including one lookup_company accepts, such as 'charges' — is a bad_request naming this tool's allowed set. |
| Name | Required | Description |
|---|---|---|
| notes | No | Caveats to surface to the user, carried over from the company report: why the list is empty, an unclassified legal form, a status that suspends filing. |
| today | Yes | The date 'next occurrence' was computed from, inclusive. Echoed back so the answer is reproducible and an agent can tell a cached answer from a fresh one. |
| country | Yes | ISO-3166-1 alpha-2, upper-case. |
| registry | Yes | Registry slug, e.g. 'brreg'. |
| deadlines | No | One entry per obligation kind, always the next occurrence, sorted by due_date. An empty list is a real answer, not an error — read `notes` for why. |
| company_id | Yes | Canonical national identifier the deadlines were computed for. |
| company_name | No | Registered name, so the caller can echo it back to a user. |
| rules_last_reviewed | Yes | The date this country's deadline rules — the statutes and their day-count arithmetic — were last checked against the law; a deadline computed long after this date should be re-verified before anyone acts on it. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already signal read-only and idempotent, and the description adds meaningful runtime behavior: empty `deadlines` is a real answer for inactive entities, `days_until` can go negative for overdue UK filings, the tool performs an upstream entity lookup first, and `rules_last_reviewed` exposes when the legal rules were last verified. No contradiction with annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is long but every block earns its place: country coverage, parameter behavior, output interpretation, error handling, and legal-source caveats are all separately and clearly organized. The one-sentence purpose is front-loaded before any country-specific detail.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the multi-country complexity, the description covers what an agent needs to invoke the tool correctly and interpret results: correct identifier shapes, country differences, how `include` alters Swedish computations, how to quote `due_date` versus `statutory_date`, the meaning of an empty `deadlines` list, error envelope behavior, and the freshness caveat from `rules_last_reviewed`. The output schema covers the remaining return structure.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema already covers 100% of parameters, but the description adds substantial semantics beyond it: `country` spells out what each jurisdiction covers, `include` explains that only 'filings' can change a computed deadline and that Norway/UK accept it without effect, and `today` is tied to reproducibility. It also clarifies that `id` is normalised from several input formats.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The first sentence names a specific verb and resource: 'Give the next occurrence of each statutory filing deadline a company faces.' It then distinguishes the tool by country and filing type, and the sibling tools (lookup_company, search_company, list_countries) are clearly different operations.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives extensive context on how to call the tool: which country codes are supported, how `today` makes results reproducible, which output fields to quote, and that `include=['filings']` changes Swedish year-end assumptions. It does not explicitly say 'use this instead of lookup_company when you need deadlines,' but the functional separation is clear enough.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
ChatGPT connector alias; other clients should prefer lookup_company plus
company_deadlines, which return the CompanyReport and DeadlineReport shapes directly.
Takes one id from search — "{COUNTRY}:{identifier}", e.g. "NO:923609016" — and
returns that company's register record and statutory filing deadlines as readable text,
both full JSON documents in metadata.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | An `id` from a `search` result: '{COUNTRY}:{identifier}', e.g. 'NO:923609016' or 'GB:00445790'. |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| url | Yes | |
| text | Yes | |
| title | Yes | |
| metadata | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover read-only, open-world, idempotent, and non-destructive behavior. The description adds that output is 'readable text' with 'full JSON documents in metadata', which is useful return-format context beyond annotations. Missing details like pagination or error behavior, but the additive value is solid.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences, zero filler. Every clause adds information: alias status, alternatives, id format, and return content. Front-loads the most important scoping detail.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For an alias tool, all necessary aspects are covered: purpose, usage guidance, parameter format, return shape, and pointers to preferred alternatives. An agent can correctly invoke it without further investigation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Single parameter id is fully described in schema with examples and format specification, and the prose description repeats the format. 100% schema coverage ensures no parameter is undocumented.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description clearly states it fetches a company record and deadlines, explicitly labels itself as a ChatGPT connector alias, and differentiates from siblings by naming lookup_company and company_deadlines as preferred alternatives. The verb and resource are specific.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly says other clients should prefer lookup_company plus company_deadlines, giving clear when-not-to-use guidance. Also explains the id comes from search, implying a workflow sequence. No ambiguity about when to use this tool.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List every national company registry this service can answer for right now, with each
one's identifier scheme, source URL, licence, supported_includes, and whether the
upstream register needs a credential (requires_api_key, api_key_env).
Call it before your first lookup in a country you have not used here, whenever a user
names a country you are unsure of, or before guessing an include value — never
hard-code a country list of your own, since it grows as modules are added. Stub modules
are hidden; only registries that actually answer are listed. No error mode.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| countries | No | One row per registry that can answer right now, sorted by country code. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover read-only, idempotent, and non-destructive traits. The description adds genuine value beyond that: stub modules are hidden so only responding registries appear, and there is 'No error mode.' No contradiction with annotations found.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two purposeful paragraphs front-loading purpose and output, then usage guidance. Slightly longer than strictly necessary, but every sentence earns its place — the usage directives and growth caveat are not filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Complete for a parameterless discovery tool. Purpose, output fields, when to invoke, behavioral caveats (stub hiding, no error mode, dynamic growth) are all present, and the output schema covers return-value documentation. Nothing an agent needs to call it correctly is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
With zero parameters and schema coverage at 100%, there is nothing the schema must document. The description compensates by detailing what the result contains (identifier scheme, source URL, licence, supported_includes, requires_api_key, api_key_env), giving the agent a clear picture of the payload it will receive.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb and resource — 'List every national company registry this service can answer for right now' — and enumerates the output fields (identifier scheme, source URL, licence, supported_includes, credential needs). This clearly differentiates it from lookup/search siblings, which operate on individual companies, not on the registry catalog itself.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Gives explicit when-to-use conditions: before first lookup in an unused country, when uncertain about a country, or before guessing an 'include' value. It also issues an imperative — 'never hard-code a country list' — which is a strong usage directive an agent can act on. No exclusions are needed given the broad list-then-lookup workflow.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Look up a company by its national identifier and get the full CompanyReport — legal form, status, address, VAT registration where the register publishes it, board and accounts duties, employees, and more.
country="NO" is the norway company lookup for the norwegian business registry:
Brønnøysundregistrene / Enhetsregisteret (brreg), by organisasjonsnummer (orgnr,
org.nr). country="GB" is the uk company lookup at Companies House, by company number
(company registration number, CRN) — "UK" is not a country code here. country="SE" is
the swedish company lookup at Bolagsverket, by organisationsnummer or a sole trader's
(enskild näringsidkare) personnummer, and by identifier only, since Bolagsverket's free
API has no name search.
include=[...] attaches seven second fetches, each with its own provenance and
null unless you ask: filings (filing history — do they file, and on time),
charges (registered mortgages and security interests), insolvency (winding-up
and administration), financials (annual accounts — turnover, operating result,
profit, balance sheet: the solvency question), lei (the GLEIF Legal Entity
Identifier), parents (direct and ultimate parent — this entity's group — from
GLEIF) and peppol (whether an e-invoice would reach them, ahead of Norway's 1
January 2027 EHF duty). The include argument explains each: what it returns, which
countries declare it, how to read its nulls.
Use it once you have the identifier — from the user, an invoice, a contract, or a
search_company hit's id. Read the returned notes before acting: it carries
caveats such as bankruptcy, dissolution, a deleted entity, an unclassified legal
form, or an attachment whose own fetch failed.
This tool does not perform sanctions, PEP or adverse-media screening, and it does not verify bank account details — it returns identity and filing data from the national register only, never a compliance clearance or a confirmed payment detail.
Errors are the {"error": {"code", "message", "hint"}} envelope this server's
instructions set out code by code (D-007); hint names the next call. A failed
attachment fetch is not one of them: the base report still comes back, that block
is left null, and notes says which attachment failed and why.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape. | |
| country | No | ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set. | NO |
| include | No | Attachment names to fetch alongside the base report; empty by default, which costs exactly one upstream request. Each is a second, independent fetch attached at that name with its own provenance, null unless you ask, and this argument is where each of the seven is explained. 'filings' (every country): what the entity has filed, and when — Companies House the whole filing history, Bolagsverket the filed annual reports, Regnskapsregisteret the filed annual accounts; the block's notes says which, and total_count how many more the register holds. 'charges' (GB): registered mortgages and other security interests, in the register's own words. 'insolvency' (GB): winding-up and administration proceedings — a members' voluntary liquidation is a *solvent* wind-up, so is_liquidation: true is not by itself evidence of distress. 'financials' (NO and SE): the register's own figures for the latest filed accounting period — turnover, operating result, profit, balance sheet, equity, liabilities, each beside its currency, never a ratio or a verdict. Norway's come in the 'filings' fetch, Sweden's out of the entity's own filed K2 annual report. A null figure in a present block means the company did not report that line; an absent block means you did not ask, the fetch failed, or — Sweden — it has filed no digital annual report, and notes says which. Britain does not declare it, so 'financials' for GB is a bad_request, never an empty block. 'lei' (every country except Sweden, whose identifier can be a natural person's): the Legal Entity Identifier GLEIF, the Global LEI Foundation, publishes — CC0 and keyless; lei: null in a present block means GLEIF holds none. 'parents' (same countries as 'lei'): the direct and ultimate parent from GLEIF's Level 2 data — the entity that consolidates this one's accounts into its group, not necessarily its majority shareholder. Where GLEIF discloses none, that side carries the entity's own stated reason as a category word such as 'NATURAL_PERSONS' — never a name, and unverified. 'peppol' (Norway): whether an e-invoice can reach the entity over the Peppol network, read live from the SML/SMP walk the way ELMA resolves it, ahead of the 1 January 2027 EHF (Peppol BIS Billing 3.0) duty. registered: null means no authoritative answer — never read it as "no"; only an NXDOMAIN or an SMP 404 earns false. Read a country's supported_includes from list_countries first: a value it does not declare is a bad_request naming what it does support, never an empty result. |
| Name | Required | Description |
|---|---|---|
| id | Yes | Canonical national identifier, digits/letters only, no spaces or dots. |
| lei | No | This entity's Legal Entity Identifier record, from GLEIF — not a national register (D-026(c), D-045(e)). `None` unless `lei` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for why). A country that declares this attachment (`CountryInfo.supported_includes`) returns a *present* block even for an entity GLEIF holds no LEI for: `LeiRecord.lei` is `None` inside it, which is a real answer, never the same as this field being absent (D-011). Declared by default by every country whose identifiers cannot be a natural person's (`Registry.universal_includes`) — Sweden does not declare it, because GLEIF is a third-party host queried by identifier in a URL query string. |
| euid | No | European Unique Identifier (EUID, Commission Implementing Regulation (EU) 2021/1042 Article 9), where the register publishes one, e.g. Finland's 'FIFPRO.0112038-9'. None for a register that does not (today: all of ours). Three traps: (1) this is not the LEI — the EUID is register-issued, mandatory in the EU and free, the LEI is voluntary, global, LOU-issued and fee-bearing; an entity may carry both, one or neither. (2) 'EUid' also names the EU Digital Identity wallet, a personal credential unrelated to company registers. (3) it is not stable across a register reorganisation, since it encodes the register of origin (e.g. France's RNE replacing the RCS in 2023). Carried verbatim from the register; never constructed from parts. |
| name | Yes | Current registered name. |
| No | Contact email as registered. | |
| notes | No | Caveats an agent should surface to the user, plain English, one per item. |
| phone | No | Contact phone as registered. |
| cached | No | True when served from our cache rather than a live fetch. |
| peppol | No | Whether this entity can be reached over the Peppol e-invoicing network. `None` unless `peppol` was passed in `include=[...]` — and, even then, `None` only if the attachment could not be built at all (see `notes` for why). A country that declares this attachment returns a **present** block even when the network could not be reached: `PeppolParticipant.registered` carries the three-state answer (`true`/`false`/`null`) and `PeppolParticipant.participant_id` is always populated, because that is the key a caller needs to ask elsewhere regardless (D-011, D-029(c)). Norway only, today (D-046(h)): the participant identifier needs a country's own ISO 6523 ICD and the answer's provenance is a different SMP per participant, neither of which generalises to `Registry.universal_includes` yet. |
| sector | No | Institutional sector description. |
| source | No | Human-readable source name, e.g. 'Enhetsregisteret (brreg.no)'. |
| status | No | Normalised lifecycle status. |
| charges | No | Registered charges (mortgages / security interests) against this entity. `None` unless `charges` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). A country that declares this attachment (`CountryInfo.supported_includes`) returns a *present* block with an empty `charges` list for an entity that genuinely has none — the two states never collapse into each other (D-011, D-042(d)). |
| country | Yes | ISO-3166-1 alpha-2, upper-case, e.g. 'NO'. |
| filings | No | What this entity has filed with its register, and when. `None` unless `filings` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). Scope differs by country because each register publishes a different subset, and the block's own `notes` says which: Companies House the whole filing history, Bolagsverket the filed annual reports, Regnskapsregisteret the filed annual accounts. A *present* block with `documents: []` means the register lists none — never the same as absent (D-011, D-042(d)). |
| license | No | Licence of the upstream data, e.g. 'NLOD 2.0'. |
| parents | No | Corporate parents from GLEIF Level 2 — accounting consolidation, not shareholding (see `ParentBlock`'s own docstring). `None` unless `parents` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for why). A country that declares this attachment returns a *present* block even for an entity GLEIF holds no LEI for at all, with `direct` and `ultimate` both `None`: GLEIF cannot hold Level 2 for an entity it has no Level 1 for, and that is the answer, not an absence (D-011). Declared by default by every country whose identifiers cannot be a natural person's (`Registry.universal_includes`, alongside `lei`) — Sweden does not declare it, for the same reason it does not declare `lei`. |
| purpose | No | Statutory purpose / objects clause, joined into one string. |
| website | No | Website as registered. |
| activity | No | Free-text description of actual activity. |
| in_group | No | Part of a corporate group. |
| registry | Yes | Registry slug, e.g. 'brreg'. |
| employees | No | Registered number of employees. None = not reported. |
| id_scheme | No | Name of the identifier scheme, e.g. 'organisasjonsnummer'. |
| is_active | No | Convenience mirror of `status == active`, so agents need no enum table. |
| parent_id | No | Identifier of the parent/owning entity, if any. |
| registers | No | Other national sub-registers this entity is or is not in, keyed by a lower-case slug, e.g. {'stiftelsesregisteret': false}. |
| confidence | No | How sure we are this record is the entity the caller meant (D-005). |
| fetched_at | No | UTC timestamp of the live fetch this record came from. |
| financials | No | Key figures from this entity's filed annual accounts. `None` unless `financials` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). Norway returns a *present* block with `periods: []` for an entity Regnskapsregisteret holds no filed accounts for. Sweden instead returns *no block at all*, plus a report-level `notes` sentence naming the reason, because Bolagsverket's digital annual-report channel holds nothing for that entity — a fact about the company, not about the country (D-042(d)(3), `tasks/T55.md`). Norway and Sweden today: Norway's arrive in the register's own open key-figures feed, Sweden's are read out of the entity's own filed annual report. Britain does not declare this attachment because the accounts of the companies that matter are filed on paper or as PDF ahead of the 1 April 2028 machine-readable mandate — a fact about the register's own population, not a parser this project has declined to write (D-043(i), D-047(f)). |
| founded_at | No | Incorporation / foundation date. |
| insolvency | No | Insolvency proceedings the register publishes against this entity. `None` unless `insolvency` was passed in `include=[...]`, or that fetch failed. A *present* block with `cases: []` means the register publishes no case, which for Companies House is the normal answer for a solvent company — and is not evidence the entity exists, since that register answers the same way for a number never issued. Read `InsolvencyCase.is_liquidation` with its own caveat: a members' voluntary liquidation is a solvent wind-up. |
| is_subunit | No | True when this record is a branch/sub-unit, not a legal entity. |
| legal_form | No | English label, e.g. 'Private limited company'. |
| source_url | No | Direct URL of the upstream record, for citation. |
| vat_number | No | VAT identifier if it differs from `id` (Norway: id + 'MVA'). |
| sector_code | No | Institutional sector code. |
| id_formatted | No | The identifier as a local would write it, e.g. '923 609 016'. |
| registered_at | No | Date first entered in the central register. |
| share_capital | No | Registered share capital. |
| status_detail | No | One sentence in English explaining the status and the flag it came from. |
| has_board_duty | No | True when this legal form must have a registered board. |
| industry_codes | No | Industry classifications, primary first. |
| postal_address | No | Postal address. |
| previous_names | No | Former registered names, newest first. |
| vat_registered | No | Registered for VAT (Norway: Merverdiavgiftsregisteret). |
| bankruptcy_date | No | Date bankruptcy was opened. |
| deregistered_at | No | Date the entity was deleted from the register. |
| legal_form_code | No | National legal-form code, e.g. 'AS', 'ASA', 'ENK'. |
| business_address | No | Visiting/registered office. |
| confidence_basis | No | Why that confidence, e.g. 'exact identifier lookup'. |
| legal_form_local | No | Local label, e.g. 'Aksjeselskap'. |
| limited_liability | No | True when owners are not personally liable for debts. |
| vat_registered_at | No | Date of VAT registration. |
| employees_reported | No | Whether the registry holds an employee figure at all (distinguishes 0 from unknown). |
| published_deadlines | No | Filing dates the upstream register publishes for this entity itself, carried verbatim. Empty for a register that publishes none — most of them. This is the input `Registry.deadlines(report, today)` needs to prefer the register's own figure over any calculation (DECISIONS.md D-018), and it is what keeps that method the pure function of (report, today) its contract promises. |
| in_business_register | No | Listed in the commercial register (Norway: Foretaksregisteret). |
| advertising_protected | No | Whether the register marks this entity as protected against direct-marketing use (Danish CVR-loven § 19 'reklamebeskyttelse', Swedish 'reklamspärr'). True: the register marks it. False: the register publishes such a flag for this entity and it is not set. None: this register publishes no such flag at all — the default, and it must never default to False, since False asserts a claim about a register that made none. When True, a country module must also append a `notes` entry containing the phrase 'direct marketing' (case-insensitive) stating the protection — that phrase is the contract this model enforces (see the validator below) — because the marking is a legal condition of passing this record's contact details on, and it must travel with them. |
| share_capital_currency | No | ISO-4217 code for `share_capital`. |
| has_annual_accounts_duty | No | True when this legal form must file annual accounts with the state. |
| last_annual_accounts_year | No | Most recent financial year for which accounts were filed. |
| business_register_registered_at | No | Date entered in the commercial/business register, where that is separate. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and openWorldHint, and the description supplements them with important non-obvious behavior: attachments are lazy second fetches that can fail without failing the base report, notes carries caveats, the error envelope format is disclosed, and the tool returns register data only. No contradiction with annotations exists.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is front-loaded with purpose, then country rules, include semantics, usage, limitations, and errors. It is long but proportionate to the multi-country, multi-attachment complexity; the slight redundancy with the schema's include parameter documentation keeps it from being maximally concise.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a complex tool with three countries, seven optional attachments, an output schema, and sibling search tools, the description covers country-specific behavior, attachment null semantics, error handling, caveats in notes, and explicit non-goals. Nothing an agent needs to call it correctly seems missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline is 3, but the description adds meaningful context beyond the schema, such as Bolagsverket having no name search, why SE lookup is identifier-only, and what the include attachments return. The schema already carries most parameter detail, so this is not a 5.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific verb and resource: look up a company by its national identifier and return the full CompanyReport. It distinguishes itself from search_company by explicitly saying to use this tool once you have an identifier, including one from a search_company hit.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives explicit when-to-use guidance: use it once you have the identifier from a user, invoice, contract, or search_company hit. It also gives when-not-to-use information by saying it does not perform sanctions, PEP, adverse-media, or bank-account verification, and it defines supported country codes while rejecting UK.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
ChatGPT connector alias; other clients should prefer search_company, which takes an
explicit country and returns the full SearchResult. One free-text query — a name, an
identifier, or either plus a country — across Norway, the United Kingdom and Sweden.
Returns {"results": [{"id", "title", "url"}]}; pass a result's id to fetch.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | A company name, a national identifier, or either plus a country. |
| Name | Required | Description |
|---|---|---|
| results | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, so the agent knows it's a safe read operation. The description adds that the query can be a name, identifier, or plus country, and that it scans three countries, and returns a specific structure. However, it doesn't mention whether results are exhaustive or if there are limitations on query length or count. With annotations covering the safety profile, a 3 is appropriate as the description adds some behavioral context but not extensive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single paragraph of two sentences, which is reasonably concise and front-loaded with the alias note and alternatives. It packs essential information without excessive verbosity. However, it could be slightly more structured with a line break for readability, but overall it earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given that the tool has a single parameter, complete schema coverage, annotations for safety, and an output schema that defines the return structure, the description is almost complete. It explains the alias relationship, the query types, the geographic coverage, and the return format. The only minor missing pieces are details about pagination or limits, but these are not critical given the simplicity of the tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents the 'query' parameter with examples and description. The description in the tool description largely reiterates the same information about the query being a name, identifier, or plus country, adding no new semantics beyond what the schema provides. Since the description does not add significant new meaning, baseline 3 is correct.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states that the tool performs a search for companies across specific countries and returns a structured result with id, title, and url. It clearly identifies the resource and action. However, it doesn't explicitly distinguish its behavior from 'search_company' beyond noting it's an alias, which is a minor gap in differentiation.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly states that 'other clients should prefer `search_company`', which takes an explicit `country` and returns full SearchResult. This provides clear guidance on when to use this tool versus the sibling, and also indicates that this alias is intended for ChatGPT connectors. The description also outlines the acceptable query formats and the geographic scope, offering strong usage context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search a national company register by name, when you have a name rather than an identifier.
country="NO" searches Brønnøysundregistrene / Enhetsregisteret (brreg) — the norway
company lookup for the norwegian business registry when the organisasjonsnummer (orgnr,
org.nr) is not yet known; country="GB" is the uk company search at Companies House,
returning each hit's company number (company registration number, CRN).
Sweden cannot be searched by name. Bolagsverket's free API has four operations and
none takes a company name, so country="SE" raises not_implemented — a fact about
the register, not a temporary gap, and it will not start working. Sweden supports
lookup by identifier only: call lookup_company with the ten-digit
organisationsnummer (or a sole trader's twelve-digit personnummer), or
validate_company_id first to check the shape for free. Bolagsverket publishes the
whole register as bulk downloadable files for callers who must search by name.
Then call lookup_company with the id of the right hit for the full report — a
search hit is deliberately thin (name, legal form, status, city) and must not be acted
on directly. Hits arrive in the register's own relevance order, so read each hit's
confidence rather than assuming the first row is best. Zero hits is not an error, and
hint says what to try next — Norwegian names are registered upper-case and often carry
an 'AS', 'ASA' or 'NUF' suffix, UK names a 'LIMITED', 'LTD', 'PLC' or 'LLP' one, worth
dropping before concluding a company does not exist.
Errors are the {"error": {"code", "message", "hint"}} envelope this server's
instructions set out code by code; hint names the next call. Call list_countries
if you are unsure a country is supported.
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | Company name to search for, free text — not an identifier. Use lookup_company once you have the id of the right hit. | |
| limit | No | Maximum hits to return, 1-100; default 10. Outside that range is a bad_request, not a silent clamp. | |
| country | No | ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set. | NO |
| Name | Required | Description |
|---|---|---|
| hint | No | What to do next, e.g. 'call lookup_company with the id of the right hit'. |
| hits | No | Best matches, best first: always sorted by `confidence` descending. Hits that score equally keep the order the upstream register returned them in. |
| query | Yes | The name that was searched for. |
| total | No | Total matches upstream, which may exceed len(hits). |
| cached | No | Served from cache. |
| country | Yes | ISO-3166-1 alpha-2, upper-case. |
| registry | Yes | Registry slug. |
| truncated | No | True when `total` exceeds the returned hits. |
| fetched_at | No | UTC timestamp of the fetch. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the annotations, the description discloses multiple non-obvious behaviors: search hits are deliberately thin and must not be acted on directly; relevance order is register-defined and confidence should be read rather than assuming the first hit; zero hits is not an error; errors use the standard envelope; and country=SE raises a permanent not_implemented. This goes well beyond the structured metadata.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is long but dense and exceptionally well structured: it starts with the core purpose, then organizes by country, then covers behavior and error handling. Every sentence carries operational relevance, and no filler is present.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has three parameters, country-specific behavior, a non-trivial error contract, and a requirement to hand off to lookup_company, the description is thorough. It explains zero-hit behavior, hit confidence, naming suffix pitfalls, and the follow-up flow, so an agent has everything needed to invoke it correctly without further research.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is already 100%, but the description adds important semantic nuance: name is free text and not an identifier, limit outside 1-100 produces bad_request rather than a silent clamp, and country must be ISO-3166-1 alpha-2 with a specific caveat that 'UK' is rejected and SE is unsupported. This is strong added meaning beyond the input schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific verb and resource: search a national company register by name, and explicitly distinguishes this from identifier lookup via lookup_company. It also separates the national registers by country, so an agent can immediately tell this tool apart from its siblings.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives clear when-to-use guidance: 'when you have a name rather than an identifier.' It also names alternatives — lookup_company, validate_company_id, list_countries — and explicitly warns that Sweden cannot be searched by name and requires lookup_company. This exceeds mere context and gives concrete routing rules.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Check whether a national company identifier is well-formed — no network call.
country="NO" checksum-checks a Norwegian organisasjonsnummer (orgnr, org.nr) for
Brønnøysundregistrene / Enhetsregisteret (brreg) — the cheap norway company lookup
pre-check for the norwegian business registry. country="GB" shape-checks and
normalises a UK company number (company registration number, CRN) for Companies House
('445790' → '00445790', 'oc303675' → 'OC303675'); a CRN has no check digit, so a GB
valid: true means the shape is right and nothing more. country="SE" shape-checks and
normalises a Swedish organisationsnummer for Bolagsverket ('556016-0680' and
'SE556016068001' both become '5560160680') and accepts a sole trader's twelve-digit
personnummer; Sweden's check digit is not enforced here (registry://rules/SE says
why), so an SE valid: true means the shape is right, reason may carry a caveat,
and the register's own verdict arrives on the lookup. It is the cheapest way to tell a
ten-digit Swedish organisationsnummer from a nine-digit Norwegian organisasjonsnummer.
Use it on user input or a spreadsheet column before spending a real lookup_company
call, since it is instant and free.
Returns a ValidationResult and never raises for a malformed identifier: valid: false
comes with reason and hint rather than a tool error — this tool answers a question,
it does not fail on bad input (D-010). A valid identifier does not mean the entity
exists; follow it with lookup_company if you need facts. The only error it raises is
unsupported_country, in the usual {"error": {"code", "message", "hint"}} envelope —
call list_countries.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape. | |
| country | No | ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set. | NO |
| Name | Required | Description |
|---|---|---|
| hint | No | What to do next when `valid` is false — the same hint the invalid_id error carries. None when valid: the next call is simply lookup. |
| input | Yes | The identifier exactly as the caller supplied it. |
| valid | Yes | True when the identifier passes this country's format and checksum. |
| reason | No | One English sentence saying why it is valid, or what failed. |
| country | Yes | ISO-3166-1 alpha-2, upper-case. |
| registry | Yes | Registry slug, e.g. 'brreg'. |
| formatted | No | The identifier as a local would write it, e.g. '923 609 016'. None when invalid. |
| id_scheme | No | Name of the identifier scheme, e.g. 'organisasjonsnummer'. |
| normalized | No | Canonical form to pass to lookup, e.g. '923609016'. None when invalid. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint, but the description adds substantial behavioral context beyond that: it never raises for malformed input (returns valid:false with reason/hint), raises only unsupported_country in the standard error envelope, and clarifies that GB 'valid:true' means shape only (no check digit) and SE does not enforce the check digit. This directly addresses what an agent would need to know about how the tool behaves.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is long (several paragraphs) but it is front-loaded with the core purpose and usage guidance. Each country-specific paragraph earns its place by explaining distinct validation rules and caveats. It is structured (per-country sections, then error handling) which aids scanning. The length is justified by the multi-registry complexity, though a more compact summary could tighten it slightly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a validation tool with multi-country rules and an output schema, the description is nearly exhaustive. It covers normalisation, checksum policies, error behavior, when to use, and follow-up actions. The existence of an output schema covers the return type, so the description doesn't need to explain that. Nothing an agent needs to call it correctly appears missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so baseline is 3, but the description adds significant meaning: it gives concrete normalisation examples for GB ('445790' → '00445790', 'oc303675' → 'OC303675') and SE ('556016-0680' and 'SE556016068001' both become '5560160680'), explicitly rejects 'UK' as a country code, and clarifies that spaces, dots, hyphens, and NO...MVA suffix are accepted. This goes well beyond the schema's brief parameter descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb+resource ('Check whether a national company identifier is well-formed') and immediately scopes it ('no network call'). It differentiates itself from siblings like lookup_company (which provides actual entity facts) and list_countries (which lists supported countries) by framing itself as a cheap pre-check. The multi-country coverage is explicit, so an agent can tell it apart without opening any other tool.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives explicit when-to-use guidance: 'Use it on user input or a spreadsheet column before spending a real lookup_company call' and notes it's the cheapest way to tell a Swedish from a Norwegian identifier. It also advises following with lookup_company when facts are needed, effectively naming the alternative and the condition that selects it. No exclusions are left to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
company_deadlines2 fields changedOutput schema / properties / rules_last_reviewedAdded value: +{
+ "description": "The date this country's deadline rules — the statutes and their day-count arithmetic — were last checked against the law; a deadline computed long after this date should be re-verified before anyone acts on it.",
+ "format": "date",
+ "title": "Rules Last Reviewed",
+ "type": "string"
+}Output schema / requiredPrevious value: -[
- "country",
- "registry",
- "company_id",
- "today"
-]New value: +[
+ "country",
+ "registry",
+ "company_id",
+ "today",
+ "rules_last_reviewed"
+]company_deadlines5 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set."Input schema / properties / country / examplesRemoved value: -[
- "NO",
- "GB",
- "SE"
-]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."New value: +"The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape."Input schema / properties / include / descriptionPrevious value: -"Optional attachment names that can change a computed deadline — a narrower set than lookup_company's include argument, which also offers attachments no date depends on. Today this is just 'filings': a second, independent upstream request for the entity's filing history, which supplies a real financial year end where one would otherwise be assumed to be 31 December. Costs one extra upstream request beyond the base lookup, only when asked; empty by default. Norway and the United Kingdom accept it too but it changes nothing for them today — their own dates already come from a published figure or a different computation. An include value this operation does not accept (including one lookup_company does, such as 'charges' or 'financials') is a bad_request naming the allowed set for this tool specifically."New value: +"Attachment names that can change a *computed* deadline — narrower than lookup_company's include. Today only 'filings': one extra upstream request for the entity's filing history, supplying a real financial year end where 31 December would otherwise be assumed. Empty by default; Norway and the United Kingdom accept it and it changes nothing for them today. Any other value — including one lookup_company accepts, such as 'charges' — is a bad_request naming this tool's allowed set."Input schema / properties / today / descriptionPrevious value: -"Date to compute deadlines from, YYYY-MM-DD. Defaults to the server's current UTC date — pass it explicitly for a reproducible answer. A value that is not YYYY-MM-DD is a bad_request naming the required format."New value: +"Date to compute deadlines from, YYYY-MM-DD; defaults to the server's current UTC date. Anything else is a bad_request naming the format."lookup_company5 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set."Input schema / properties / country / examplesRemoved value: -[
- "NO",
- "GB",
- "SE"
-]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."New value: +"The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape."Input schema / properties / include / descriptionPrevious value: -"Optional attachment names to fetch alongside the base report. Each is a second, independent fetch attached at that same name on the result, with its own provenance, and null unless you ask for it. Seven exist today. 'filings' (every country): what the entity has actually filed, and when. 'charges' (United Kingdom only): mortgages and other security interests. 'insolvency' (United Kingdom only): winding-up and administration proceedings. 'financials' (Norway and Sweden): the register's own financial figures — turnover, operating result, profit, a balance sheet, and more — for the latest filed accounting period, answering whether a supplier looks solvent; Norway's arrive in the same fetch as 'filings', Sweden's are read out of the entity's own filed annual report. 'lei' (every country except Sweden): the Legal Entity Identifier GLEIF publishes for the entity, CC0-licensed and keyless. 'parents' (the same countries as 'lei'): the corporate parent GLEIF's Level 2 data discloses, direct and ultimate, or the entity's own stated reason — a category word such as 'NATURAL_PERSONS', never a name — when it discloses none. 'peppol' (Norway only): whether the entity can receive an e-invoice over the Peppol network, read live from the SML/SMP walk ahead of the 1 January 2027 e-invoicing duty. Empty by default, which costs exactly one upstream request. Ask for one when the base report is not enough to answer the question in front of you: 'filings' answers whether they file on time, 'charges' whether assets are already pledged, 'insolvency' whether they are being wound up, 'financials' what the numbers say, 'lei' and 'parents' who GLEIF says this entity and its group are, 'peppol' whether an e-invoice would reach them. Call list_countries and read a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result."New value: +"Attachment names to fetch alongside the base report; empty by default, which costs exactly one upstream request. Each is a second, independent fetch attached at that name with its own provenance, null unless you ask, and this argument is where each of the seven is explained. 'filings' (every country): what the entity has filed, and when — Companies House the whole filing history, Bolagsverket the filed annual reports, Regnskapsregisteret the filed annual accounts; the block's notes says which, and total_count how many more the register holds. 'charges' (GB): registered mortgages and other security interests, in the register's own words. 'insolvency' (GB): winding-up and administration proceedings — a members' voluntary liquidation is a *solvent* wind-up, so is_liquidation: true is not by itself evidence of distress. 'financials' (NO and SE): the register's own figures for the latest filed accounting period — turnover, operating result, profit, balance sheet, equity, liabilities, each beside its currency, never a ratio or a verdict. Norway's come in the 'filings' fetch, Sweden's out of the entity's own filed K2 annual report. A null figure in a present block means the company did not report that line; an absent block means you did not ask, the fetch failed, or — Sweden — it has filed no digital annual report, and notes says which. Britain does not declare it, so 'financials' for GB is a bad_request, never an empty block. 'lei' (every country except Sweden, whose identifier can be a natural person's): the Legal Entity Identifier GLEIF, the Global LEI Foundation, publishes — CC0 and keyless; lei: null in a present block means GLEIF holds none. 'parents' (same countries as 'lei'): the direct and ultimate parent from GLEIF's Level 2 data — the entity that consolidates this one's accounts into its group, not necessarily its majority shareholder. Where GLEIF discloses none, that side carries the entity's own stated reason as a category word such as 'NATURAL_PERSONS' — never a name, and unverified. 'peppol' (Norway): whether an e-invoice can reach the entity over the Peppol network, read live from the SML/SMP walk the way ELMA resolves it, ahead of the 1 January 2027 EHF (Peppol BIS Billing 3.0) duty. registered: null means no authoritative answer — never read it as \"no\"; only an NXDOMAIN or an SMP 404 earns false. Read a country's supported_includes from list_countries first: a value it does not declare is a bad_request naming what it does support, never an empty result."Input schema / properties / include / examplesPrevious value: -[
- [
- "filings"
- ],
- [
- "charges",
- "insolvency"
- ],
- [
- "financials"
- ],
- [
- "lei"
- ],
- [
- "parents"
- ],
- [
- "peppol"
- ],
- []
-]New value: +[
+ [
+ "filings"
+ ],
+ [
+ "charges",
+ "insolvency"
+ ],
+ [
+ "financials",
+ "lei"
+ ]
+]search1 field changedInput schema / properties / query / descriptionPrevious value: -"What to look for: a company name, a national identifier, or a name plus a country, e.g. 'Equinor', '923609016', 'Tesco United Kingdom'."New value: +"A company name, a national identifier, or either plus a country."search_company3 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set."Input schema / properties / country / examplesRemoved value: -[
- "NO",
- "GB",
- "SE"
-]Input schema / properties / limit / descriptionPrevious value: -"Maximum hits to return. 1-100, default 10; a value outside that range is a bad_request, not a silent clamp."New value: +"Maximum hits to return, 1-100; default 10. Outside that range is a bad_request, not a silent clamp."validate_company_id3 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 — NO Norway, GB United Kingdom, SE Sweden. UK is not a country code here and is rejected. Call list_countries for the live set."Input schema / properties / country / examplesRemoved value: -[
- "NO",
- "GB",
- "SE"
-]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."New value: +"The company's national identifier, normalised for you — a Norwegian organisasjonsnummer (orgnr), a Companies House company number (CRN), or a Swedish organisationsnummer or personnummer. Spaces, dots, hyphens, a NO...MVA suffix and a short CRN are accepted; list_countries gives each country's exact shape."company_deadlines4 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."Input schema / properties / country / examplesPrevious value: -[
- "NO",
- "GB"
-]New value: +[
+ "NO",
+ "GB",
+ "SE"
+]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you."New value: +"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."Input schema / properties / includeAdded value: +{
+ "default": [],
+ "description": "Optional attachment names that can change a computed deadline — a narrower set than lookup_company's include argument, which also offers attachments no date depends on. Today this is just 'filings': a second, independent upstream request for the entity's filing history, which supplies a real financial year end where one would otherwise be assumed to be 31 December. Costs one extra upstream request beyond the base lookup, only when asked; empty by default. Norway and the United Kingdom accept it too but it changes nothing for them today — their own dates already come from a published figure or a different computation. An include value this operation does not accept (including one lookup_company does, such as 'charges' or 'financials') is a bad_request naming the allowed set for this tool specifically.",
+ "examples": [
+ [
+ "filings"
+ ],
+ []
+ ],
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+}lookup_company12 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."Input schema / properties / country / examplesPrevious value: -[
- "NO",
- "GB"
-]New value: +[
+ "NO",
+ "GB",
+ "SE"
+]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you."New value: +"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."Input schema / properties / include / descriptionPrevious value: -"Optional attachment names to fetch alongside the base report. Each is a second, independent fetch attached at that same name on the result, with its own provenance, and null unless you ask for it. Three exist today: 'filings' (what the entity has actually filed, and when — every country), 'charges' (mortgages and other security interests, United Kingdom only) and 'insolvency' (winding-up and administration proceedings, United Kingdom only). Empty by default, which costs exactly one upstream request. Ask for one when the base report is not enough to answer the question in front of you: 'filings' answers whether they file on time, 'charges' whether assets are already pledged, 'insolvency' whether they are being wound up. Call list_countries and read a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result."New value: +"Optional attachment names to fetch alongside the base report. Each is a second, independent fetch attached at that same name on the result, with its own provenance, and null unless you ask for it. Seven exist today. 'filings' (every country): what the entity has actually filed, and when. 'charges' (United Kingdom only): mortgages and other security interests. 'insolvency' (United Kingdom only): winding-up and administration proceedings. 'financials' (Norway and Sweden): the register's own financial figures — turnover, operating result, profit, a balance sheet, and more — for the latest filed accounting period, answering whether a supplier looks solvent; Norway's arrive in the same fetch as 'filings', Sweden's are read out of the entity's own filed annual report. 'lei' (every country except Sweden): the Legal Entity Identifier GLEIF publishes for the entity, CC0-licensed and keyless. 'parents' (the same countries as 'lei'): the corporate parent GLEIF's Level 2 data discloses, direct and ultimate, or the entity's own stated reason — a category word such as 'NATURAL_PERSONS', never a name — when it discloses none. 'peppol' (Norway only): whether the entity can receive an e-invoice over the Peppol network, read live from the SML/SMP walk ahead of the 1 January 2027 e-invoicing duty. Empty by default, which costs exactly one upstream request. Ask for one when the base report is not enough to answer the question in front of you: 'filings' answers whether they file on time, 'charges' whether assets are already pledged, 'insolvency' whether they are being wound up, 'financials' what the numbers say, 'lei' and 'parents' who GLEIF says this entity and its group are, 'peppol' whether an e-invoice would reach them. Call list_countries and read a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result."Input schema / properties / include / examplesPrevious value: -[
- [
- "filings"
- ],
- [
- "charges",
- "insolvency"
- ],
- []
-]New value: +[
+ [
+ "filings"
+ ],
+ [
+ "charges",
+ "insolvency"
+ ],
+ [
+ "financials"
+ ],
+ [
+ "lei"
+ ],
+ [
+ "parents"
+ ],
+ [
+ "peppol"
+ ],
+ []
+]Output schema / properties / charges / anyOfPrevious value: -[
- {
- "additionalProperties": false,
- "description": "Registered charges for one entity — an `include=[\"charges\"]` attachment\n(D-042(g), D-045(a)), never a plain field on `CompanyReport` (D-041(c)):\nit is a second round trip with its own moment, its own cache state and its\nown failure mode, so it carries its own `SourceRef` rather than reusing\nthe report's.\n\n**The count fields are ruled by D-045(a)**, not by D-042(h). It strikes\n`outstanding_count` — `total_count - satisfied_count` is our arithmetic\nwearing a register figure's name, and on a shared model it would mean\n\"outstanding\" for a register with no partially-satisfied state and \"not\nfully satisfied\" for one that has it (D-011) — and adds the register's\nown `part_satisfied_count` in its place. **This class has not yet been\nreconciled with that ruling**; the task that does so is named in D-045(a)\nand is due before the next deploy.\n\nTwo-level nullability is the point of this shape (D-026(c), D-041(c),\nD-042(d)(3)): `CompanyReport.charges` is `None` when `charges` was not in\n`include`, or when the fetch failed (`Registry.lookup_with` appends a\n`notes` sentence on the report saying which). Once *present*, this block\ncarries `charges: []` for an entity the register confirms has none —\nthat case must never collapse into the absent case, and must never be\n`not_found` (D-011).",
- "properties": {
- "charges": {
- "description": "Sorted newest first (a country module's own tie-break rule).",
- "items": {
- "additionalProperties": false,
- "description": "One registered charge (a mortgage or other security interest) against\nan entity — one row of a `ChargeBlock`.\n\nField names are country-neutral (D-042(g)): GB is the first filler\n(Companies House `/company/{n}/charges`, `registries/gb/__init__.py`),\nand any future filler (e.g. Norway's Løsøreregisteret, once it opens a\npublic API) maps onto this same shape rather than getting one of its own.\n\n**The field list is ruled by D-045(a)**, not by D-042(h) — which rules\n`FiledDocument` and no charge shape at all. D-045(a) accepts these names,\ncorrects `created_on`'s description (it is the date the charge instrument\nwas created, not a record timestamp), and adds `contains_fixed_charge` and\n`contains_negative_pledge` beside `contains_floating_charge`, all three of\nwhich Companies House emits **only when true** — so an absent flag means\nthe register did not mark this instrument, never that it lacks one\n(D-011). **This class has not yet been reconciled with that ruling**; the\ntask that does so is named in D-045(a) and is due before the next deploy.",
- "properties": {
- "assets_charged": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own free-text description of what is charged.",
- "title": "Assets Charged"
- },
- "charge_id": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own opaque handle for this charge; not fetchable through this API. `None` when the register has no such handle for an older filing — honestly absent, not guessed.",
- "title": "Charge Id"
- },
- "charge_number": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's sequence number for this charge, within this company.",
- "title": "Charge Number"
- },
- "classification": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "What kind of instrument this is, as the register describes it.",
- "title": "Classification"
- },
- "contains_floating_charge": {
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Whether the register marks this instrument as including a floating charge.",
- "title": "Contains Floating Charge"
- },
- "created_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was created.",
- "title": "Created On"
- },
- "delivered_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was delivered to the register for registration.",
- "title": "Delivered On"
- },
- "is_outstanding": {
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Derived from `status` by membership of a country module's own committed table of status words it has actually observed on the wire. `None` when `status` is absent or is a word not yet in that table — never guessed, never `False` by default (D-025(d), D-011).",
- "title": "Is Outstanding"
- },
- "obligations_secured": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own free-text description of what the charge secures.",
- "title": "Obligations Secured"
- },
- "parties_entitled": {
- "description": "Names exactly as the register publishes them for the party or parties the charge is entitled to (typically a bank, an insurer or a trustee company; occasionally a natural person, e.g. a director lending to their own company). This is a term of the company's own instrument, not a person record: it is never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040). It is the one place in this product a natural person's name can appear, and it is deliberately not named the register's own `persons_entitled` — that name asserts a natural person; this one does not.",
- "items": {
- "type": "string"
- },
- "title": "Parties Entitled",
- "type": "array"
- },
- "satisfied_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was satisfied, if it has been.",
- "title": "Satisfied On"
- },
- "status": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own word, verbatim, e.g. \"outstanding\", \"fully-satisfied\" — national vocabulary lives here, in the value, never in a field name (D-042(g)). See `is_outstanding` for the country-neutral derived flag.",
- "title": "Status"
- }
- },
- "title": "Charge",
- "type": "object"
- },
- "title": "Charges",
- "type": "array"
- },
- "notes": {
- "description": "Plain-English caveats about this block, e.g. truncation when `total_count` exceeds `len(charges)`.",
- "items": {
- "type": "string"
- },
- "title": "Notes",
- "type": "array"
- },
- "outstanding_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Derived as `total_count - satisfied_count` when the register publishes both as whole-company figures; a register that also tracks a distinct partially-satisfied state folds it in here as not-fully-satisfied. `None` when the register does not publish enough to derive it.",
- "title": "Outstanding Count"
- },
- "provenance": {
- "additionalProperties": false,
- "description": "Where, when and under what licence this block was fetched.",
- "properties": {
- "cached": {
- "default": false,
- "description": "True when this attachment was served from cache rather than a live fetch.",
- "title": "Cached",
- "type": "boolean"
- },
- "fetched_at": {
- "anyOf": [
- {
- "format": "date-time",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "UTC timestamp of the live fetch this attachment came from.",
- "title": "Fetched At"
- },
- "license": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
- "title": "License"
- },
- "source": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
- "title": "Source"
- },
- "source_url": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Direct URL of the upstream record, for citation.",
- "title": "Source Url"
- }
- },
- "title": "SourceRef",
- "type": "object"
- },
- "satisfied_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own whole-company count of satisfied charges, verbatim.",
- "title": "Satisfied Count"
- },
- "total_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own count of charges for this company, which may exceed `len(charges)` — see `notes` for a truncation disclosure when it does.",
- "title": "Total Count"
- }
- },
- "required": [
- "provenance"
- ],
- "title": "ChargeBlock",
- "type": "object"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "additionalProperties": false,
+ "description": "Registered charges for one entity — an `include=[\"charges\"]` attachment\n(D-042(g), D-045(a)), never a plain field on `CompanyReport` (D-041(c)):\nit is a second round trip with its own moment, its own cache state and its\nown failure mode, so it carries its own `SourceRef` rather than reusing\nthe report's.\n\n**The count fields are ruled by D-045(a)**, not by D-042(h). It strikes\n`outstanding_count` — `total_count - satisfied_count` was our arithmetic\nwearing a register figure's name, and on a shared model it would have\nmeant \"outstanding\" for a register with no partially-satisfied state and\n\"not fully satisfied\" for one that has it (D-011) — and adds the\nregister's own `part_satisfied_count` in its place: a whole-company\nfigure Companies House publishes directly, `0` in every fixture observed\nso far, and not derived from anything else on this block.\n\nTwo-level nullability is the point of this shape (D-026(c), D-041(c),\nD-042(d)(3)): `CompanyReport.charges` is `None` when `charges` was not in\n`include`, or when the fetch failed (`Registry.lookup_with` appends a\n`notes` sentence on the report saying which). Once *present*, this block\ncarries `charges: []` for an entity the register confirms has none —\nthat case must never collapse into the absent case, and must never be\n`not_found` (D-011).",
+ "properties": {
+ "charges": {
+ "description": "Sorted newest first (a country module's own tie-break rule).",
+ "items": {
+ "additionalProperties": false,
+ "description": "One registered charge (a mortgage or other security interest) against\nan entity — one row of a `ChargeBlock`.\n\nField names are country-neutral (D-042(g)): GB is the first filler\n(Companies House `/company/{n}/charges`, `registries/gb/__init__.py`),\nand any future filler (e.g. Norway's Løsøreregisteret, once it opens a\npublic API) maps onto this same shape rather than getting one of its own.\n\n**The field list is ruled by D-045(a)**, not by D-042(h) — which rules\n`FiledDocument` and no charge shape at all. D-045(a) accepts the names\nthis class started with, corrects the `created_on` / `delivered_on` /\n`satisfied_on` descriptions, and adds `contains_fixed_charge` and\n`contains_negative_pledge` beside `contains_floating_charge`, all three of\nwhich Companies House emits **only when true** — so an absent flag means\nthe register did not mark this instrument, never that it lacks one, and\nnever `False` (D-011). It also adds `assets_charged_type` and\n`obligations_secured_type`, the register's own category token for each\nfree-text field (S-series finding 7, D-042(e)(3)'s `persons_entitled`\ntreatment reused twice over).",
+ "properties": {
+ "assets_charged": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what is charged, relayed verbatim and uncapped. It may contain particulars of property, account details or a natural person's name (e.g. a guarantor): never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040) — the same binding `parties_entitled` carries. See `assets_charged_type` for the register's own category token this text is filed under.",
+ "title": "Assets Charged"
+ },
+ "assets_charged_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own category token for `assets_charged`, verbatim: `short-particulars` or `brief-description` are the two seen on Companies House's `particulars.type` — two different kinds of text under one field name, which this token disambiguates. National vocabulary lives here, in the value, never in the field name (D-042(g)).",
+ "title": "Assets Charged Type"
+ },
+ "charge_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this charge; not fetchable through this API. `None` on an older filing that predates the register assigning one — 19 of 110 items observed, all pre-2013 — honestly absent, not guessed.",
+ "title": "Charge Id"
+ },
+ "charge_number": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own sequence number for this charge, scoped to this company only: not unique across companies, and not a lookup key.",
+ "title": "Charge Number"
+ },
+ "classification": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What kind of instrument this is, in the register's own prose — not a code, and not the same field as `FiledDocument.type_code`.",
+ "title": "Classification"
+ },
+ "contains_fixed_charge": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether Companies House marks this instrument as including a fixed charge. The register emits this key **only when `True`**: absence means the register did not mark this instrument, never that it lacks one, and must never be read or stored as `False` (D-011). Norway's Løsøreregisteret and Sweden's företagsinteckningar use different taxonomies for security interests, so a future filler for either may leave this `None` on every item it fills, forever — the same shape as `FiledDocument.days_from_fee_point`.",
+ "title": "Contains Fixed Charge"
+ },
+ "contains_floating_charge": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether Companies House marks this instrument as including a floating charge. The register emits this key **only when `True`**: absence means the register did not mark this instrument, never that it lacks one, and must never be read or stored as `False` (D-011). Norway's Løsøreregisteret and Sweden's företagsinteckningar use different taxonomies for security interests, so a future filler for either may leave this `None` on every item it fills, forever — the same shape as `FiledDocument.days_from_fee_point`.",
+ "title": "Contains Floating Charge"
+ },
+ "contains_negative_pledge": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether Companies House marks this instrument as including a negative pledge — a covenant restricting further charges over the same assets, and, on its own, a fact a lender changes behaviour on. The register emits this key **only when `True`**: absence means the register did not mark this instrument, never that it lacks one, and must never be read or stored as `False` (D-011). Norway's Løsøreregisteret and Sweden's företagsinteckningar use different taxonomies for security interests, so a future filler for either may leave this `None` on every item it fills, forever — the same shape as `FiledDocument.days_from_fee_point`.",
+ "title": "Contains Negative Pledge"
+ },
+ "created_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The date the *charge instrument* was created — not a record timestamp. See `delivered_on` for the date it reached the register; the gap between the two is the Companies Act 2006 s.859A window a charge must be delivered within to be registered at all.",
+ "title": "Created On"
+ },
+ "delivered_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own receipt date: when the charge was delivered to the register for registration, which is not the same date as `created_on`. The interval between them is the Companies Act 2006 s.859A window a charge must be delivered within to be registered at all.",
+ "title": "Delivered On"
+ },
+ "is_outstanding": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Derived from `status` by membership of a country module's own committed table of status words it has actually observed on the wire. `None` when `status` is absent or is a word not yet in that table — never guessed, never `False` by default (D-025(d), D-011).",
+ "title": "Is Outstanding"
+ },
+ "obligations_secured": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what the charge secures, relayed verbatim and uncapped, under the same binding as `assets_charged` (D-028(1), D-040): never a lookup key, never indexed, never searchable, never reaches a log line. See `obligations_secured_type` for the register's own category token — on every observed item that carried one (19 of 19) it was `amount-secured`, never `obligations-secured`, so this field's name is a category the token disambiguates, not a description every value matches.",
+ "title": "Obligations Secured"
+ },
+ "obligations_secured_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own category token for `obligations_secured`, verbatim from Companies House's `secured_details.type`. Observed as `amount-secured` on 19 of 19 items that carried a token at all — `obligations_secured` itself has never once carried an `obligations-secured` value, which is exactly why this token, not the field's name, is the category. National vocabulary lives here, in the value, never in the field name (D-042(g)).",
+ "title": "Obligations Secured Type"
+ },
+ "parties_entitled": {
+ "description": "Names exactly as the register publishes them for the party or parties the charge is entitled to (typically a bank, an insurer or a trustee company; occasionally a natural person, e.g. a director lending to their own company). This is a term of the company's own instrument, not a person record: it is never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040) — the same binding `assets_charged` and `obligations_secured` carry, because free prose describing charged property can also name a guarantor or a charged dwelling. It is deliberately not named the register's own `persons_entitled` — that name asserts a natural person; this one does not.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Parties Entitled",
+ "type": "array"
+ },
+ "satisfied_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The date the register recorded this charge as satisfied, exactly as published. `None` while outstanding — see `status` for the register's own word and `is_outstanding` for the derived flag.",
+ "title": "Satisfied On"
+ },
+ "status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word, verbatim, e.g. \"outstanding\", \"fully-satisfied\" — national vocabulary lives here, in the value, never in a field name (D-042(g)). See `is_outstanding` for the country-neutral derived flag.",
+ "title": "Status"
+ }
+ },
+ "title": "Charge",
+ "type": "object"
+ },
+ "title": "Charges",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block: truncation when `total_count` exceeds `len(charges)`, and — whenever any charge on this page carries free text in `assets_charged` or `obligations_secured` — a disclosure that the text is the register's own prose, relayed verbatim and not parsed, and may name a natural person or carry an account identifier.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "part_satisfied_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own whole-company count of charges it marks partially satisfied, identical in kind to `total_count` and `satisfied_count`. `None` when the register does not publish it. Observed as `0` in every fixture this project has seen — this project has never observed a non-zero value — and the count is the register's own; it is not derived from anything else on this block.",
+ "title": "Part Satisfied Count"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "satisfied_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own whole-company count of satisfied charges, verbatim.",
+ "title": "Satisfied Count"
+ },
+ "total_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own count of charges for this company, which may exceed `len(charges)` — see `notes` for a truncation disclosure when it does.",
+ "title": "Total Count"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "ChargeBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+]Output schema / properties / filings / anyOfPrevious value: -[
- {
- "additionalProperties": false,
- "description": "What one entity has filed with its national register — an\n``include=[\"filings\"]`` attachment (D-041(d), D-042), never a plain field\non :class:`CompanyReport` (D-041(c)): it is a second round trip with its\nown moment, its own cache state and its own failure mode, so it carries its\nown :class:`SourceRef` rather than reusing the report's.\n\nAll three live countries declare it, and each answers a differently-scoped\nquestion its register actually supports — Companies House returns the whole\nfiling history, Bolagsverket the filed annual reports, Regnskapsregisteret\nthe filed annual accounts. `notes` says which, in words, on every block.\n\nTwo-level nullability is the contract (D-011, D-026(c), D-041(c)):\n**absent** means \"you did not ask, or the fetch failed\" — `lookup_with`\nappends one `notes` sentence to the report saying which — while **present\nwith `documents: []`** means \"the register lists no filings for this\nentity\", a real and useful answer about a counterparty that must never be\nrendered as an absence.",
- "properties": {
- "documents": {
- "description": "One page of the register's own filing history, newest first by `filed_at`. Never paginated further; when the register holds more, `total_count` says how many and `notes` says so in words. Empty means the register lists none — not that we could not look.",
- "items": {
- "additionalProperties": false,
- "description": "One filing a national register publishes for one entity — one row of a\n:class:`FilingHistory`.\n\nThe shape ``DECISIONS.md`` D-041(d) ruled and D-042(h) widened, and it is\ncountry-neutral by construction rather than by intent: Britain, Sweden and\nNorway each built this model independently behind their own seam, and all\nthree arrived field-for-field at this one. National vocabulary lives in the\n*values* (`category`, `type_code`, `description_code`), never in a field\nname (D-042(g)).\n\nEvery field is nullable and every `None` means the same thing: **the\nregister does not publish it** (D-011). It never means zero, never means\n\"no\", and is never filled by derivation — a register that does not publish\na period start gets `None`, not a start inferred by subtracting twelve\nmonths from the end (D-009).",
- "properties": {
- "category": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own category for this filing, verbatim and never translated — \"accounts\", \"mortgage\", \"confirmation-statement\", \"gazette\" and some twenty more in Britain alone, and none of these lists is closed. It is the field `kind` is derived from.",
- "title": "Category"
- },
- "days_from_fee_point": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Signed days from a named late-fee datum to `filed_at`, **only where the register itself publishes such a datum for that period**. Negative is early. `None` is the common answer and means the datum does not exist in the data, not that the arithmetic was skipped: Sweden fills it because årsredovisningslagen 8 kap. 6 § names one datum for every company, while Companies House publishes only the *next* period's due date and nothing per-period historical, so there is nothing to measure a past filing against. Deriving one from the statutory rule would require guessing a period length, a first-accounts variant and any shortening the register has not disclosed, then presenting the result as the register's own — the invented figure D-009 forbids.",
- "title": "Days From Fee Point"
- },
- "description_code": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own description-template key, verbatim and **never resolved into prose**. This is the key and not the sentence on purpose, and the reason is the whole design of this block: Companies House resolves these templates from a `description_values` object, 97 templates interpolate an officer's name and 26 a person with significant control's, so the resolved sentence is personal data while the key is not. **The key says what happened; only the values say who** (D-042(e)(1), D-028).",
- "title": "Description Code"
- },
- "document_id": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own opaque handle for this filing, relayed verbatim and never interpreted. **Not fetchable through this API**: the filed document itself lives behind a separate host, which is a second upstream with its own provenance and out of scope for this block (D-041(c)). It is the key a support case with the register can name.",
- "title": "Document Id"
- },
- "file_format": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "What the register holds the document as, where it says. `None` where the filing-history endpoint publishes no format — Companies House keeps the media type on its separate document host, a second fetch this block does not make.",
- "title": "File Format"
- },
- "filed_at": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the register recorded this filing, verbatim. This is the field that makes the block answer *does this company file on time*, and it is the sort key for `FilingHistory.documents`: newest first.",
- "title": "Filed At"
- },
- "kind": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The `Deadline.kind` slug this filing discharges, or `None` when it discharges none. This is the one field that is *derived* rather than relayed, and it is derived only by a committed per-country table of category words actually observed on the wire. A filing whose category is outside that table gets `None` rather than an invented slug (D-009): a filing that discharges no deadline this product publishes says so honestly.",
- "title": "Kind"
- },
- "period_end": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The reporting period's last day, exactly as the register published it. Beware what the period belongs to: on an annual-accounts filing it is the date the accounts were made up to, but a register may publish a made-up date on other filing kinds too — a British confirmation statement carries one, and it is not a financial year end. Read it together with `kind`. `None` on the great majority of filings, which have no reporting period at all.",
- "title": "Period End"
- },
- "period_start": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The reporting period's first day, as published. `None` wherever the register publishes no counterpart to `period_end` — deriving one would assert a period length the register never stated, and a first, shortened or extended accounting period is lawful and common (D-009). Norway's Regnskapsregisteret publishes `regnskapsperiode: {fraDato, tilDato}` and fills both ends; Companies House publishes only the end.",
- "title": "Period Start"
- },
- "type_code": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own form code for this filing, verbatim: \"AA\", \"CS01\", \"AP01\", \"MR01\" and older forms such as \"288a\" and \"363s\" in Britain — 100 distinct codes across 1876 items observed live.",
- "title": "Type Code"
- }
- },
- "title": "FiledDocument",
- "type": "object"
- },
- "title": "Documents",
- "type": "array"
- },
- "financial_year_end": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The latest reporting period among this entity's filed **annual accounts** (`kind == \"annual_accounts\"`), carried verbatim — never a synthesised month-day, and never taken from a filing of another kind that happens to carry a made-up date of its own. It is the latest *period*, not the period of the latest *filing*, because a register may accept a later filing that amends an earlier year and that would otherwise roll this date backwards. It is **evidence of** the entity's accounting reference date, not a statement of it. `None` when this page holds no annual-accounts filing with a reporting period, including when older accounts exist further back than the page reaches.",
- "title": "Financial Year End"
- },
- "notes": {
- "description": "Plain-English caveats about this block: which subset of filings this register publishes, truncation when `total_count` exceeds `len(documents)`, and which empty state an empty `documents` is.",
- "items": {
- "type": "string"
- },
- "title": "Notes",
- "type": "array"
- },
- "provenance": {
- "additionalProperties": false,
- "description": "Where, when and under what licence this block was fetched.",
- "properties": {
- "cached": {
- "default": false,
- "description": "True when this attachment was served from cache rather than a live fetch.",
- "title": "Cached",
- "type": "boolean"
- },
- "fetched_at": {
- "anyOf": [
- {
- "format": "date-time",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "UTC timestamp of the live fetch this attachment came from.",
- "title": "Fetched At"
- },
- "license": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
- "title": "License"
- },
- "source": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
- "title": "Source"
- },
- "source_url": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Direct URL of the upstream record, for citation.",
- "title": "Source Url"
- }
- },
- "title": "SourceRef",
- "type": "object"
- },
- "total_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own count of filings for this entity, which may greatly exceed `len(documents)` — 8371 against a 25-row page, for one company observed live. **`None` means the register published no count**, not zero, and for Companies House it additionally distinguishes a real zero from a number whose filing history the register cannot serve at all: that endpoint returns `0` for both, and relaying the second as a zero would assert something the register never said (D-011). `notes` names which case it was.",
- "title": "Total Count"
- }
- },
- "required": [
- "provenance"
- ],
- "title": "FilingHistory",
- "type": "object"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "additionalProperties": false,
+ "description": "What one entity has filed with its national register — an\n``include=[\"filings\"]`` attachment (D-041(d), D-042), never a plain field\non :class:`CompanyReport` (D-041(c)): it is a second round trip with its\nown moment, its own cache state and its own failure mode, so it carries its\nown :class:`SourceRef` rather than reusing the report's.\n\nAll three live countries declare it, and each answers a differently-scoped\nquestion its register actually supports — Companies House returns the whole\nfiling history, Bolagsverket the filed annual reports, Regnskapsregisteret\nthe filed annual accounts. `notes` says which, in words, on every block.\n\nTwo-level nullability is the contract (D-011, D-026(c), D-041(c)):\n**absent** means \"you did not ask, or the fetch failed\" — `lookup_with`\nappends one `notes` sentence to the report saying which — while **present\nwith `documents: []`** means \"the register lists no filings for this\nentity\", a real and useful answer about a counterparty that must never be\nrendered as an absence.",
+ "properties": {
+ "documents": {
+ "description": "One page of the register's own filing history, newest first by the register's own period or filing date: Britain sorts by `filed_at`; Sweden by `period_end` then `filed_at`; Norway by `period_end` then `period_start`. Never paginated further; when the register holds more, `total_count` says how many and `notes` says so in words. Empty means the register lists none — not that we could not look.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One filing a national register publishes for one entity — one row of a\n:class:`FilingHistory`.\n\nThe shape ``DECISIONS.md`` D-041(d) ruled and D-042(h) widened, and it is\ncountry-neutral by construction rather than by intent: Britain, Sweden and\nNorway each built this model independently behind their own seam, and all\nthree arrived field-for-field at this one. National vocabulary lives in the\n*values* (`category`, `type_code`, `description_code`), never in a field\nname (D-042(g)).\n\nEvery field is nullable and every `None` means the same thing: **the\nregister does not publish it** (D-011). It never means zero, never means\n\"no\", and is never filled by derivation — a register that does not publish\na period start gets `None`, not a start inferred by subtracting twelve\nmonths from the end (D-009).",
+ "properties": {
+ "category": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own category for this filing, verbatim and never translated. It is the field `kind` is derived from. Twenty-two words observed live in Britain, and the list is not closed: \"accounts\", \"capital\", \"officers\", \"mortgage\", \"confirmation-statement\", \"annual-return\", \"resolution\", \"gazette\", \"incorporation\", \"address\", \"insolvency\", \"dissolution\", \"change-of-name\", \"persons-with-significant-control\", \"auditors\", \"miscellaneous\", \"historical\", \"restoration\", \"document-replacement\", \"change-of-constitution\", \"return\" and \"other\".",
+ "title": "Category"
+ },
+ "days_from_fee_point": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Signed days from a named late-fee datum to `filed_at`, **only where the register itself publishes such a datum for that period**. Negative is early. `None` is the common answer and means the datum does not exist in the data, not that the arithmetic was skipped. Sweden's Bolagsverket fills it: årsredovisningslagen 8 kap. 6 § starts a förseningsavgift of 7 500 kr (15 000 kr for a public company) at that datum. It is **not** the company's own filing deadline — ÅRL 8 kap. 3 § instead requires filing within one month of the general meeting that adopts the accounts — and a nine-month variant of 8 kap. 6 § cannot be excluded, because the dataset does not identify which companies it applies to. Companies House publishes only the *next* period's due date — `accounts.next_accounts.due_on` and `confirmation_statement.next_due` on the company profile — and no per-period historical due date at all, so there is nothing to measure a past filing against without guessing a 9-month or 6-month period and presenting the guess as the register's own — the invented figure D-009 forbids.",
+ "title": "Days From Fee Point"
+ },
+ "description_code": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own description-template key, verbatim and **never resolved into prose**. This is the key and not the sentence on purpose, and the reason is the whole design of this block: Companies House resolves these templates from a `description_values` object, 97 templates interpolate an officer's name and 26 a person with significant control's, so the resolved sentence is personal data while the key is not. **The key says what happened; only the values say who** (D-042(e)(1), D-028).",
+ "title": "Description Code"
+ },
+ "document_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this filing, relayed verbatim and never interpreted. **Not fetchable through this API**: the filed document itself lives behind a separate host, which is a second upstream with its own provenance and out of scope for this block (D-041(c)). It is the key a support case with the register can name. Norway's payload also carries an integer `id`, an internal row identifier, which is not relayed.",
+ "title": "Document Id"
+ },
+ "file_format": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What the register holds the document as, where it says. `None` where the filing-history endpoint publishes no format: Companies House's filing-history endpoint publishes only a page count and a `paper_filed` marker, no media type — the media type itself lives on a separate document host, a second fetch this block does not make.",
+ "title": "File Format"
+ },
+ "filed_at": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the register recorded this filing, verbatim. This is the field that makes the block answer *does this company file on time*, and it is the sort key for `FilingHistory.documents`: newest first.",
+ "title": "Filed At"
+ },
+ "kind": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The `Deadline.kind` slug this filing discharges, or `None` when it discharges none. This is the one field that is *derived* rather than relayed, and it is derived only by a committed per-country table of category words actually observed on the wire. A filing whose category is outside that table gets `None` rather than an invented slug (D-009): a filing that discharges no deadline this product publishes says so honestly.",
+ "title": "Kind"
+ },
+ "period_end": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's last day, exactly as the register published it. Beware what the period belongs to: on an annual-accounts filing it is the date the accounts were made up to, but a register may publish a made-up date on other filing kinds too — a British confirmation statement carries one, and it is not a financial year end. Read it together with `kind`. `None` on the great majority of filings, which have no reporting period at all.",
+ "title": "Period End"
+ },
+ "period_start": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's first day, as published. `None` wherever the register publishes no counterpart to `period_end` — deriving one would assert a period length the register never stated, and a first, shortened or extended accounting period is lawful and common (D-009). Norway's Regnskapsregisteret publishes `regnskapsperiode: {fraDato, tilDato}` and fills both ends; Companies House publishes only the end; so does Sweden's Bolagsverket, whose bokföringslagen 3 kap. 3 § permits an 18-month first or final period — exactly the period length a subtracted twelve months would falsely assert.",
+ "title": "Period Start"
+ },
+ "type_code": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own form code for this filing, verbatim: \"AA\", \"CS01\", \"AP01\", \"MR01\" and older forms such as \"288a\" and \"363s\" in Britain — 100 distinct codes across 1876 items observed live.",
+ "title": "Type Code"
+ }
+ },
+ "title": "FiledDocument",
+ "type": "object"
+ },
+ "title": "Documents",
+ "type": "array"
+ },
+ "financial_year_end": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The latest reporting period among this entity's filed **annual accounts** (`kind == \"annual_accounts\"`), carried verbatim — never a synthesised month-day, and never taken from a filing of another kind that happens to carry a made-up date of its own. It is the latest *period*, not the period of the latest *filing*, because a register may accept a later filing that amends an earlier year and that would otherwise roll this date backwards. It is **evidence of** the entity's accounting reference date, not a statement of it. `None` when this page holds no annual-accounts filing with a reporting period, including when older accounts exist further back than the page reaches.",
+ "title": "Financial Year End"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block: which subset of filings this register publishes, truncation when `total_count` exceeds `len(documents)`, and which empty state an empty `documents` is.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "total_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own count of filings for this entity, which may greatly exceed `len(documents)` — 8371 against a 25-row page, for one company observed live. **`None` means the register published no count**, not zero, and for Companies House it additionally distinguishes a real zero from a number whose filing history the register cannot serve at all: that endpoint returns `0` for both, and relaying the second as a zero would assert something the register never said (D-011). `notes` names which case it was.",
+ "title": "Total Count"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "FilingHistory",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+]Output schema / properties / financialsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Key figures from an entity's filed annual accounts — an\n``include=[\"financials\"]`` attachment (DECISIONS.md D-043), never a plain\nfield on :class:`CompanyReport` (D-041(c)): it is a second round trip with\nits own moment, its own cache state and its own failure mode, so it\ncarries its own :class:`SourceRef` rather than reusing the report's.\n\n**A second block, not a wider `FiledDocument`** (D-043(b)): \"did they file\non time\" and \"what do the numbers say\" are two different questions, and\nfolding nineteen numeric fields onto `FiledDocument` would collapse two\nmeanings into one `None` — \"Britain does not publish this\" and \"this\nNorwegian company did not report this line\" — which D-011 forbids.\n\nNorway and Sweden fill this block today, from two different sources:\nNorway's figures arrive in Regnskapsregisteret's own open key-figures\nfeed, the same fetch as `filings`; Sweden's are read out of the entity's\nown filed annual report (the K2 inline-XBRL document Bolagsverket's\ndocument API serves), a second request that shares its document-list\ndiscovery step with `filings` but is not the same fetch (DECISIONS.md\nD-047(f)). Britain does not fill this block, and that is the register's\nown population, not a scope decision this project made: the accounts of\nthe companies that matter are filed on paper or as PDF, the\nmachine-readable (iXBRL) mandate is 1 April 2028 with a\nprofit-and-loss publication opt-out for small and micro companies, and\nno British filing sampled carried the balance-sheet totals this block\nrelays (DECISIONS.md D-043(i), as amended by D-047(f)).\n`include=[\"financials\"]` on a country that does not declare it is\n`bad_request`, never a silently empty block.\n\nTwo-level nullability is the point of the shape (D-011, D-026(c),\nD-041(c), D-042(d)(3)): once *present*, this block carries `periods: []`\nfor an entity Regnskapsregisteret holds no filed accounts for — that\nstate must never collapse into the absent state and must never be\n``not_found``. **No field on this block or on `FinancialPeriod` is a\nderived ratio, indicator or verdict** — DECISIONS.md D-043(e) rules out an\nequity ratio, a current ratio, a net-debt figure, a working-capital\nfigure and every similar field on three independent grounds, the\nstrongest being that the register relays filings whose own totals do not\nreconcile on 27 of 358 observed filings. The one comparison this project\nmakes is a `notes` sentence, never a number: see `notes` below.",
+ "properties": {
+ "notes": {
+ "description": "Plain-English caveats about this block. Unconditional on any non-empty block: that figures are denominated in the stated currency and framework and are not comparable across companies or borders without regard to both, and that this is the latest filed period rather than a history. Conditional: a reconciliation note when `total_assets` and `total_equity_and_liabilities` disagree, a non-NOK currency note, and one note each for `small_entity`, `audit_exempt`, `liquidation_basis` and an observed `unaudited` (DECISIONS.md D-043(e),(g)).",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "periods": {
+ "description": "Filed accounting periods, sorted newest first by `period_end`. Both registers carry exactly one today, for different reasons: Regnskapsregisteret publishes only one — the endpoint takes no year argument and holds no history — while Bolagsverket lists every filed annual report but only the most recent is parsed into this block, a bounded-cost choice rather than a register limit (DECISIONS.md D-047(f)). Either way this is a latest-figures block, not a trend; `notes` says so on every non-empty block, and the rest of a Swedish entity's filed reports are in `include=[\"filings\"]`. The list exists for a register that publishes more than one.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One filed accounting period's key figures, mapped a second time\nalongside :class:`~registry_mcp.core.models.FiledDocument` for the same\nfiling (D-043(h), D-047(f)). For Norway the two come from the same\nfetch; for Sweden `financials` shares `filings`' document-list discovery\nstep but then reads the filed document itself, a further fetch `filings`\nnever makes. Either way, `document_id` and `period_end` are the join\nkeys a caller uses to line this period up with its sibling\n`FiledDocument`.\n\n`currency` is the one field in this whole block with no default\n(DECISIONS.md D-043(d)): a figure separated from its currency is not\npartially wrong, it is meaningless, and this model makes constructing one\na `pydantic.ValidationError` rather than a silently-`None` currency. A\nperiod the register published with no `valuta` is not carried at all —\nthe mapper skips it and says why in `FinancialSummary.notes` — which is\nsafe because `valuta` was present on 573 of 573 payloads this project has\nread.",
+ "properties": {
+ "accounting_framework": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The accounting framework this period was prepared under, verbatim from `regnkapsprinsipper.regnskapsregler` — observed values include 'regnskapslovenAlminneligRegler', 'IFRS' and 'forenkletAnvendelseIFRS'. Two Norwegian companies' figures are not necessarily on the same basis; no field here converts between them (DECISIONS.md D-043(d)).",
+ "title": "Accounting Framework"
+ },
+ "audit_exempt": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether the company has resolved to opt out of audit under aksjeloven § 7-6, from `revisjon.fravalgRevisjon` — lawful below that section's thresholds and True on 65 of 358 observed filings (18%). The consequence a credit decision must weigh: no independent auditor checked these figures.",
+ "title": "Audit Exempt"
+ },
+ "balance_sheet": {
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Stocks at the period's last instant — Regnskapsregisteret's\n``eiendeler`` and ``egenkapitalGjeld``, the other of the two sub-objects on\na :class:`FinancialPeriod` (D-043(c)). See :class:`IncomeStatement` for why\nthe two are separate models rather than one flat one.\n\nEvery field is `None` or a whole-unit figure in :attr:`FinancialPeriod.currency`\n— never both `None` and zero at once, and never inferred from the other\n(D-043(f)). **No field here is a ratio or a verdict** — an equity ratio, a\ncurrent ratio and every similar derived figure are declined by D-043(e): the\nregister's own `total_assets` and `total_equity_and_liabilities` disagree\non 27 of 358 filings (7.5%), so a ratio built from this block would be a\nratio of two numbers the register itself does not vouch for jointly. The\none comparison this project makes is a `notes` sentence, never a number —\nsee :class:`FinancialSummary`.",
+ "properties": {
+ "current_assets": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Current assets at period end (`sumOmloepsmidler`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Current Assets"
+ },
+ "current_liabilities": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Current liabilities at period end (`sumKortsiktigGjeld`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Current Liabilities"
+ },
+ "equity": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total equity at period end (`sumEgenkapital`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Equity"
+ },
+ "fixed_assets": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Fixed assets at period end (`sumAnleggsmidler`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Fixed Assets"
+ },
+ "liabilities": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total liabilities at period end (`sumGjeld`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). Carried exactly as the register states it, including negative: DECISIONS.md D-043(e) records real filings with a negative `sumGjeld` (e.g. -108,837), which is a filing the register relayed without validating, not a company fact this field corrects.",
+ "title": "Liabilities"
+ },
+ "non_current_liabilities": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Non-current liabilities at period end (`sumLangsiktigGjeld`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Non Current Liabilities"
+ },
+ "paid_in_equity": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Paid-in equity at period end (`sumInnskuttEgenkaptial` — the register's own spelling, not a typo in this field's description). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Paid In Equity"
+ },
+ "retained_equity": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Retained equity at period end (`sumOpptjentEgenkapital`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Retained Equity"
+ },
+ "total_assets": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total assets at period end (`sumEiendeler`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). One of only three fields present on every one of 573 observed filings. Compare with `total_equity_and_liabilities` (DECISIONS.md D-043(e)): the two disagree on 27 of 358 filings, and this block's own `notes` names the gap when they do; neither figure is edited, reconciled or dropped.",
+ "title": "Total Assets"
+ },
+ "total_equity_and_liabilities": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total equity and liabilities at period end (`sumEgenkapitalGjeld`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). One of only three fields present on every one of 573 observed filings. See `total_assets` for the reconciliation note the two together can trigger.",
+ "title": "Total Equity And Liabilities"
+ }
+ },
+ "title": "BalanceSheet",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Stocks at this period's last instant. `None` when the register published no line in this statement at all for this filing; otherwise present with whichever lines it published, each individually nullable (DECISIONS.md D-043(f))."
+ },
+ "consolidated": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this filing is a consolidated (group) statement, derived from `scope` by a committed table of words this module has actually observed on the wire — today only `{'SELSKAP': False}`. A word outside that table, including an implied-but-unobserved 'KONSERN', gets `None`, never `False` (DECISIONS.md D-011, D-025(d)): this field never guesses.",
+ "title": "Consolidated"
+ },
+ "currency": {
+ "description": "ISO-4217-shaped currency code, verbatim from `valuta`. **Required — this field has no default, and a period the register published with no currency is not constructed at all** (DECISIONS.md D-043(d)): a figure without its currency is not partially wrong, it is meaningless, and this model makes that state unrepresentable rather than merely discouraged. 12 of 358 observed Norwegian filings are not in kroner (USD, EUR, SEK, DKK), so two *Norwegian* companies can be incomparable without either crossing a border. Values are whole units of this currency; scale (thousands, millions) is not recorded because the register does not publish one, and the figures are exact integers that are not significant to that precision.",
+ "title": "Currency",
+ "type": "string"
+ },
+ "document_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this filing, verbatim from `journalnr` — the same handle the sibling `FiledDocument.document_id` on `filings` carries for the same filing, and the join key between the two blocks (DECISIONS.md D-043(h)). Not fetchable through this API.",
+ "title": "Document Id"
+ },
+ "income_statement": {
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Flows over one reporting period — Regnskapsregisteret's\n``resultatregnskapResultat``, one of two sub-objects on a\n:class:`FinancialPeriod` (D-043(c)). Nested apart from :class:`BalanceSheet`\non purpose: revenue is a flow over a span, not a stock at an instant, and a\ncaller who mixes the two time semantics makes precisely the error this\nsplit exists to prevent.\n\nEvery field is `None` or a whole-unit figure in :attr:`FinancialPeriod.currency`\n— never both `None` and zero at once, and never inferred from the other\n(D-043(f)); see `FinancialPeriod.currency` for what the unit is and why it\nis required.",
+ "properties": {
+ "financial_costs": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Financial costs for the period (`sumFinanskostnad`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Financial Costs"
+ },
+ "financial_income": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Financial income for the period (`sumFinansinntekter`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Financial Income"
+ },
+ "net_financial_items": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Net financial items for the period (`nettoFinans`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Net Financial Items"
+ },
+ "operating_costs": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total operating costs for the period (`sumDriftskostnad`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Operating Costs"
+ },
+ "operating_result": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Operating result for the period (`driftsresultat`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Operating Result"
+ },
+ "profit_before_tax": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Ordinary result before tax (`ordinaertResultatFoerSkattekostnad`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)).",
+ "title": "Profit Before Tax"
+ },
+ "profit_for_period": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Profit or loss for the period (`aarsresultat`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). One of only three fields present on every one of 573 observed filings.",
+ "title": "Profit For Period"
+ },
+ "revenue": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Turnover for the period (`sumDriftsinntekter`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). Absent on 51 of 358 observed filings (14%) — not rare.",
+ "title": "Revenue"
+ },
+ "total_comprehensive_income": {
+ "anyOf": [
+ {
+ "type": "number"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Total comprehensive income for the period (`totalresultat`). `None` means the register did not publish this line for this filing — never zero. The register is inconsistent about when it states an explicit zero for an absent figure versus omitting the line entirely, so this field's absence is not evidence the true value is zero, and a stated zero is not evidence the register omits the line elsewhere (DECISIONS.md D-043(f)). Absent on 209 of 358 observed filings (58%) — the register's own line, not this project's omission.",
+ "title": "Total Comprehensive Income"
+ }
+ },
+ "title": "IncomeStatement",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Flows for this period. `None` when the register published no line in this statement at all for this filing; otherwise present with whichever lines it published, each individually nullable (DECISIONS.md D-043(f))."
+ },
+ "liquidation_basis": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this filing is an *avviklingsregnskap* under aksjeloven § 16-10 — a winding-up account prepared on a realisation rather than a going-concern basis, over a final stub period — from `avviklingsregnskap`. Rare (3 of 215 entities the register marks `underAvvikling`) and, when true, real: read `FinancialSummary.notes` for the caveat this triggers. It does not restate the winding-up itself, which `CompanyReport.status` already carries.",
+ "title": "Liquidation Basis"
+ },
+ "period_end": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's last day, as published (`regnskapsperiode.tilDato`). Equal to the sibling `FiledDocument.period_end` for this filing on the same report (DECISIONS.md D-043(h)).",
+ "title": "Period End"
+ },
+ "period_start": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's first day, as published (`regnskapsperiode.fraDato`). Real, published data — never derived by subtracting twelve months from `period_end`, because a first or final period may be shorter or longer (DECISIONS.md D-009).",
+ "title": "Period Start"
+ },
+ "scope": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What this filing covers, the register's own word, verbatim from `regnskapstype` — 'SELSKAP' (company accounts) is the only value observed in 573 payloads; 'KONSERN' (consolidated) is implied by the vocabulary but was never seen. See `consolidated` for the country-neutral derived flag.",
+ "title": "Scope"
+ },
+ "small_entity": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this filing was prepared under the reduced-disclosure regime for a *lite foretak* (regnskapsloven § 1-6), from `regnkapsprinsipper.smaaForetak`. **Not a distress signal** — True on 315 of 358 observed filings, the majority case — it is a disclosure caveat: fewer figures exist, and those that do were prepared under rules that permit simplification.",
+ "title": "Small Entity"
+ },
+ "unaudited": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Relayed uninverted from `revisjon.ikkeRevidertAarsregnskap`, which its own name claims means these accounts were not audited. **`True` was never observed** in 573 sampled payloads, including every filing by a company that had opted out of audit under `audit_exempt` — so this flag's semantics are unverified: do not read a `False` here as an assertion that the accounts were audited, and do not read this field as more reliable than `audit_exempt` (DECISIONS.md D-043(g)).",
+ "title": "Unaudited"
+ }
+ },
+ "required": [
+ "currency"
+ ],
+ "title": "FinancialPeriod",
+ "type": "object"
+ },
+ "title": "Periods",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched. For Norway, **identical in all five fields to the sibling `filings` block's `provenance` when both are requested together**: they are the same upstream fetch, not two (DECISIONS.md D-043(h)). For Sweden the two blocks' `provenance` are **not** identical: `financials` shares `filings`' document-list discovery fetch to decide what to fetch, but then makes its own further request for the document itself, and this field describes that further request, not the shared list (DECISIONS.md D-047(f)).",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "FinancialSummary",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Key figures from this entity's filed annual accounts. `None` unless `financials` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). Norway returns a *present* block with `periods: []` for an entity Regnskapsregisteret holds no filed accounts for. Sweden instead returns *no block at all*, plus a report-level `notes` sentence naming the reason, because Bolagsverket's digital annual-report channel holds nothing for that entity — a fact about the company, not about the country (D-042(d)(3), `tasks/T55.md`). Norway and Sweden today: Norway's arrive in the register's own open key-figures feed, Sweden's are read out of the entity's own filed annual report. Britain does not declare this attachment because the accounts of the companies that matter are filed on paper or as PDF ahead of the 1 April 2028 machine-readable mandate — a fact about the register's own population, not a parser this project has declined to write (D-043(i), D-047(f))."
+}Output schema / properties / insolvency / anyOfPrevious value: -[
- {
- "additionalProperties": false,
- "description": "Insolvency proceedings a register publishes against one entity — an\n``include=[\"insolvency\"]`` attachment (D-042), never a plain field on\n:class:`CompanyReport` (D-041(c)), and carrying its own :class:`SourceRef`.\n\nTwo-level nullability is the point of the shape (D-011, D-026(c),\nD-041(c), D-042(d)(3)): once *present*, this block carries ``cases: []``\nfor an entity the register publishes no insolvency case for — that state\nmust never collapse into the absent state and must never be ``not_found``.\nCompanies House's 404 here is the *normal* answer for a solvent company and\nis byte-identical to its answer for a number that was never issued, so it\nsays nothing about whether the entity exists; ``notes`` therefore\ndistinguishes \"the register holds no insolvency resource here\" from \"the\nresource exists and is empty\" instead of flattening both into silence.",
- "properties": {
- "cases": {
- "description": "Every insolvency case the register publishes for this entity — the whole history, not a page, where the register's endpoint is unpaginated. Sorted newest first by the case's most recent event date, then by `case_number` descending; cases the register gives no date for sort last.",
- "items": {
- "additionalProperties": false,
- "description": "One insolvency case a register publishes against one entity.\n\n**No practitioner particular can land here.** A register commonly publishes\neach appointed practitioner's name and postal address alongside the case;\nD-042(e)(2) bars relaying them in the first tranche, so this model has no\nfield for them and no country mapper reads the key. Adding them later is a\ndecision with its own entry in ``DECISIONS.md``, inheriting D-028's four\npreconditions in full.",
- "properties": {
- "case_number": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own identifier for this case, verbatim. For Companies House this is a per-company sequence number rendered as a string (\"1\", \"2\", … up to \"31\" in the live sample) and is **not** a court reference — it identifies the case only within this entity. Kept as a string because another register's case identifier need not be numeric.",
- "title": "Case Number"
- },
- "case_type": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own word for the kind of procedure, verbatim — national vocabulary in the value (D-042(g)). Ten words observed live in Britain, among them \"compulsory-liquidation\", \"creditors-voluntary-liquidation\", \"members-voluntary-liquidation\" and \"in-administration\". See `is_liquidation` for the country-neutral derived flag.",
- "title": "Case Type"
- },
- "events": {
- "description": "The register's own dated steps in this case, newest first. Frequently empty — 172 of the 1,485 live British cases carried no date at all, most of them old receiverships — and an empty list means the register publishes no date for this case, never that nothing happened.",
- "items": {
- "additionalProperties": false,
- "description": "One dated step in an insolvency case, as the register itself records it.\n\nThese are the register's own events, not this service's interpretation of\nthem: D-042(e)(2) rules that case type, case number and *these* dated\nevents carry the entire distress signal a pre-contract check needs.",
- "properties": {
- "event_type": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own word for what happened, verbatim — national vocabulary lives here, in the value, never in a field name (D-042(g)). Thirteen words have been observed live in Britain, from \"petitioned-on\" and \"wound-up-on\" to \"declaration-solvent-on\" and \"dissolved-on\". A word outside the observed set is still relayed verbatim: this field is never filtered, only reported.",
- "title": "Event Type"
- },
- "occurred_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The date the register gives for this event.",
- "title": "Occurred On"
- }
- },
- "title": "InsolvencyEvent",
- "type": "object"
- },
- "title": "Events",
- "type": "array"
- },
- "is_liquidation": {
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Whether this procedure is a winding-up — the country-neutral question behind the national word in `case_type`. Derived by membership of a committed table of words the country module has actually observed on the wire. `None` when `case_type` is absent or is a word not yet in that table — never guessed, never `False` by default (D-011, D-025(d)). **`True` does not mean insolvent**: a members' voluntary liquidation is a *solvent* winding-up, begun by a declaration of solvency, and 56 of the 1,485 live British cases behind this table were exactly that. Read it as 'the entity is being wound up', not as 'the entity cannot pay'.",
- "title": "Is Liquidation"
- },
- "note_codes": {
- "description": "The register's own note **codes** for this case, verbatim and never resolved into prose — the same treatment D-042(e)(1) gives a filing's `description_code`. Companies House declares this field an unbounded `array[string]`, so it is the one place in that payload a name could hide; codes are therefore relayed through an allow-list of observed codes, and an unrecognised one is dropped and disclosed in the block's `notes` rather than passed through.",
- "items": {
- "type": "string"
- },
- "title": "Note Codes",
- "type": "array"
- }
- },
- "title": "InsolvencyCase",
- "type": "object"
- },
- "title": "Cases",
- "type": "array"
- },
- "notes": {
- "description": "Plain-English caveats about this block: which of the register's two empty states this is, that practitioner particulars exist upstream and are deliberately not relayed, and any note code withheld by the allow-list.",
- "items": {
- "type": "string"
- },
- "title": "Notes",
- "type": "array"
- },
- "provenance": {
- "additionalProperties": false,
- "description": "Where, when and under what licence this block was fetched.",
- "properties": {
- "cached": {
- "default": false,
- "description": "True when this attachment was served from cache rather than a live fetch.",
- "title": "Cached",
- "type": "boolean"
- },
- "fetched_at": {
- "anyOf": [
- {
- "format": "date-time",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "UTC timestamp of the live fetch this attachment came from.",
- "title": "Fetched At"
- },
- "license": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
- "title": "License"
- },
- "source": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
- "title": "Source"
- },
- "source_url": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Direct URL of the upstream record, for citation.",
- "title": "Source Url"
- }
- },
- "title": "SourceRef",
- "type": "object"
- },
- "statuses": {
- "description": "The register's own entity-level insolvency status words, verbatim — national vocabulary in values (D-042(g)). Eight observed live in Britain, among them \"in-administration\", \"liquidation\" and \"voluntary-arrangement\". An **empty list means the register publishes no such word for this entity**, which is not the same as 'not currently insolvent': about one in ten companies whose Companies House status is itself an insolvency status still has no word here. No yes/no flag is derived from this field for exactly that reason (D-011).",
- "items": {
- "type": "string"
- },
- "title": "Statuses",
- "type": "array"
- }
- },
- "required": [
- "provenance"
- ],
- "title": "InsolvencyBlock",
- "type": "object"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "additionalProperties": false,
+ "description": "Insolvency proceedings a register publishes against one entity — an\n``include=[\"insolvency\"]`` attachment (D-042), never a plain field on\n:class:`CompanyReport` (D-041(c)), and carrying its own :class:`SourceRef`.\n\nTwo-level nullability is the point of the shape (D-011, D-026(c),\nD-041(c), D-042(d)(3)): once *present*, this block carries ``cases: []``\nfor an entity the register publishes no insolvency case for — that state\nmust never collapse into the absent state and must never be ``not_found``.\nCompanies House's 404 here is the *normal* answer for a solvent company and\nis byte-identical to its answer for a number that was never issued, so it\nsays nothing about whether the entity exists; ``notes`` therefore\ndistinguishes \"the register holds no insolvency resource here\" from \"the\nresource exists and is empty\" instead of flattening both into silence.",
+ "properties": {
+ "cases": {
+ "description": "Every insolvency case the register publishes for this entity — the whole history, not a page, where the register's endpoint is unpaginated. Sorted newest first by the case's most recent event date, then by `case_number` descending; cases the register gives no date for sort last.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One insolvency case a register publishes against one entity.\n\n**No practitioner particular can land here.** A register commonly publishes\neach appointed practitioner's name and postal address alongside the case;\nD-042(e)(2) bars relaying them in the first tranche, so this model has no\nfield for them and no country mapper reads the key. Adding them later is a\ndecision with its own entry in ``DECISIONS.md``, inheriting D-028's four\npreconditions in full.",
+ "properties": {
+ "case_number": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own identifier for this case, verbatim. For Companies House this is a per-company sequence number rendered as a string (\"1\", \"2\", … up to \"31\" in the live sample) and is **not** a court reference — it identifies the case only within this entity. Kept as a string because another register's case identifier need not be numeric.",
+ "title": "Case Number"
+ },
+ "case_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word for the kind of procedure, verbatim — national vocabulary in the value (D-042(g)). Ten words observed live in Britain, among them \"compulsory-liquidation\", \"creditors-voluntary-liquidation\", \"members-voluntary-liquidation\" and \"in-administration\". See `is_liquidation` for the country-neutral derived flag.",
+ "title": "Case Type"
+ },
+ "events": {
+ "description": "The register's own dated steps in this case, newest first. Frequently empty — 172 of the 1,485 live British cases carried no date at all, most of them old receiverships — and an empty list means the register publishes no date for this case, never that nothing happened.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One dated step in an insolvency case, as the register itself records it.\n\nThese are the register's own events, not this service's interpretation of\nthem: D-042(e)(2) rules that case type, case number and *these* dated\nevents carry the entire distress signal a pre-contract check needs.",
+ "properties": {
+ "event_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word for what happened, verbatim — national vocabulary lives here, in the value, never in a field name (D-042(g)). Thirteen words have been observed live in Britain: \"administration-started-on\", \"administration-ended-on\", \"administration-discharged-on\", \"instrumented-on\", \"petitioned-on\", \"wound-up-on\", \"concluded-winding-up-on\", \"voluntary-arrangement-started-on\", \"voluntary-arrangement-ended-on\", \"moratorium-started-on\", \"declaration-solvent-on\", \"due-to-be-dissolved-on\" and \"dissolved-on\" (the last of which Companies House's own published enumeration omits). A word outside the observed set is still relayed verbatim: this field is never filtered, only reported.",
+ "title": "Event Type"
+ },
+ "occurred_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The date the register gives for this event.",
+ "title": "Occurred On"
+ }
+ },
+ "title": "InsolvencyEvent",
+ "type": "object"
+ },
+ "title": "Events",
+ "type": "array"
+ },
+ "is_liquidation": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this procedure is a winding-up — the country-neutral question behind the national word in `case_type`. Derived by membership of a committed table of words the country module has actually observed on the wire. `None` when `case_type` is absent or is a word not yet in that table — never guessed, never `False` by default (D-011, D-025(d)). **`True` does not mean insolvent**: a members' voluntary liquidation is a *solvent* winding-up, begun by a declaration of solvency, and 56 of the 1,485 live British cases behind this table were exactly that. Read it as 'the entity is being wound up', not as 'the entity cannot pay'.",
+ "title": "Is Liquidation"
+ },
+ "note_codes": {
+ "description": "The register's own note **codes** for this case, verbatim and never resolved into prose — the same treatment D-042(e)(1) gives a filing's `description_code`. Only one code has ever been observed live: \"scottish-insolvency-info\", which means the Accountant in Bankruptcy's Register of Insolvencies holds further detail this API does not. Companies House declares this field an unbounded `array[string]`, so it is the one place in that payload a name could hide; codes are therefore relayed through an allow-list of observed codes, and an unrecognised one is dropped and disclosed in the block's `notes` rather than passed through.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Note Codes",
+ "type": "array"
+ }
+ },
+ "title": "InsolvencyCase",
+ "type": "object"
+ },
+ "title": "Cases",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block: which of the register's two empty states this is, that practitioner particulars exist upstream and are deliberately not relayed, and any note code withheld by the allow-list.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "statuses": {
+ "description": "The register's own entity-level insolvency status words, verbatim — national vocabulary in values (D-042(g)). Eight observed live in Britain: \"in-administration\", \"liquidation\", \"receivership\", \"receiver-manager\", \"administrative-receiver\", \"administration-order\", \"voluntary-arrangement\" and \"live-receiver-manager-on-at-least-one-charge\". An **empty list means the register publishes no such word for this entity**, which is not the same as 'not currently insolvent': about one in ten companies whose Companies House status is itself an insolvency status still has no word here. No yes/no flag is derived from this field for exactly that reason (D-011).",
+ "items": {
+ "type": "string"
+ },
+ "title": "Statuses",
+ "type": "array"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "InsolvencyBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+]Output schema / properties / leiAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "The Legal Entity Identifier GLEIF (the Global LEI Foundation) publishes\nfor one entity — the ``include=[\"lei\"]`` attachment, D-026(c)'s shape,\nunamended by D-045(e). Unlike every other attachment, this upstream is\nnot any one country's own register: GLEIF publishes every jurisdiction\nfrom one endpoint, under one CC0 licence, with one TTL, so it is the\nfirst attachment every country declares by default\n(``Registry.universal_includes``, D-045(e)) rather than something a\ncountry module opts into.\n\nThe two-level nullability is D-026(c)'s and D-011's, restated here: an\n**absent** ``CompanyReport.lei`` means the attachment was not requested,\nor the fetch failed (a ``notes`` sentence on the report says which); a\n**present** block with ``lei=None`` means GLEIF holds no LEI for this\nentity, which is a real and useful answer about a counterparty and must\nnever be rendered as though nothing were known.",
+ "properties": {
+ "legal_name": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The legal name as GLEIF publishes it, carried verbatim. May differ from `CompanyReport.name`, because the two are two registers' opinions recorded at two different moments — never reconciled against it and never used to correct the company record (D-018: say which source said what).",
+ "title": "Legal Name"
+ },
+ "lei": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The 20-character Legal Entity Identifier GLEIF publishes for this entity. `None` *inside a present block* means GLEIF holds no LEI for it — a real and useful answer about a counterparty, and not the same as this block being absent (D-026(c), D-011). The LEI is **not** the EUID — see `CompanyReport.euid`'s own description for that distinction rather than restating it here.",
+ "title": "Lei"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block. Always names the exact string sent as GLEIF's `entity.registeredAs` filter, so a `lei: null` answer is legible rather than silent, and the register-authority code GLEIF cites for this entity. If GLEIF returned more than one record for this registration number, a sentence discloses how many and each LEI, rather than silently picking one.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched. `source` names GLEIF, `license` is 'CC0 1.0', and `source_url` is this record's own GLEIF URL. Never implies endorsement and never describes registry-mcp as a GLEIF service — GLEIF's anti-impersonation clause sits outside its data licence (D-026(c)).",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "registration_status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "GLEIF's own `registration.status` for this LEI record, verbatim: 'ISSUED', 'LAPSED', … (national-vocabulary-in-values, D-042(g)). A 'LAPSED' LEI means the entity stopped renewing its registration and is **not** evidence of insolvency or inactivity — Carillion plc ('03782379') and Lehman Brothers International (Europe) ('02538254') both read `entity.status: ACTIVE` beside `registration.status: LAPSED`. This field carries the *registration's* status; GLEIF's `entity.status` is not carried at all, because it is a claim about the company made by neither the company's own register nor us.",
+ "title": "Registration Status"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "LeiRecord",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "This entity's Legal Entity Identifier record, from GLEIF — not a national register (D-026(c), D-045(e)). `None` unless `lei` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for why). A country that declares this attachment (`CountryInfo.supported_includes`) returns a *present* block even for an entity GLEIF holds no LEI for: `LeiRecord.lei` is `None` inside it, which is a real answer, never the same as this field being absent (D-011). Declared by default by every country whose identifiers cannot be a natural person's (`Registry.universal_includes`) — Sweden does not declare it, because GLEIF is a third-party host queried by identifier in a URL query string."
+}Output schema / properties / parentsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Corporate parents from GLEIF Level 2 — the `include=[\"parents\"]`\nattachment (D-047(a)).\n\n**Three-level nullability**, restated for this block: an **absent**\n`CompanyReport.parents` means `parents` was not requested, or the fetch\nfailed (`notes` on the report says which). A **present** block whose\n`direct` and `ultimate` are both `None` means GLEIF holds **no LEI at\nall** for this entity — there is no Level 2 to hold either, and that is\nitself the answer, not an absence. A **present** block with `direct`\nand/or `ultimate` populated is the ordinary case, whether populated side\ndiscloses a parent or records why it did not.\n\n`direct` and `ultimate` are two independent sides that can disagree in\nkind — one disclosed, the other excepted, or two different parents.\n**Never collapse them and never derive one from the other**: in a\n600-record Swedish sample, 9 records disclosed a direct parent and 10\ndisclosed an ultimate one, so at least one record took different paths\non its two sides.\n\nThe sentence the whole block hangs on: **GLEIF Level 2 reports the\n*accounting consolidating* parent — the entity that consolidates this\nentity's accounts (wire word `IS_DIRECTLY_CONSOLIDATED_BY` /\n`IS_ULTIMATELY_CONSOLIDATED_BY`) — which is not the same as the majority\nshareholder, and neither implies the other.**",
+ "properties": {
+ "direct": {
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "One side (`direct` or `ultimate`) of a `ParentBlock` — either GLEIF\ndiscloses a corporate parent for this side, or the entity's own filer\nexplained why it did not (``include=[\"parents\"]``, D-047(a)).\n\n**Exactly one of `lei` and `reporting_exception` is populated — never\nboth, never neither** (enforced below, D-011: a link that discloses a\nparent's LEI and a link that records why no parent was disclosed are two\ndifferent facts and must not collapse into one).\n\nGLEIF Level 2 reports the *accounting consolidating* parent — the wire\nword is `IS_DIRECTLY_CONSOLIDATED_BY` / `IS_ULTIMATELY_CONSOLIDATED_BY` —\nwhich is **not** the same as the majority shareholder, and neither\nimplies the other.",
+ "properties": {
+ "corroboration_level": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "GLEIF's own corroboration level for the *relationship* record, verbatim: `FULLY_CORROBORATED`, `PARTIALLY_CORROBORATED`, `ENTITY_SUPPLIED_ONLY`. Measured over 31 disclosed relationships: 10 / 11 / 10 — roughly a third of all disclosed parent links are the filer's own assertion that nobody has checked, which is the reason this field costs a second request. `None` on the exception side, where there is no relationship record.",
+ "title": "Corroboration Level"
+ },
+ "jurisdiction": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's own registered jurisdiction, ISO-3166-1 alpha-2, as GLEIF publishes it. May differ from this entity's own `CompanyReport.country`.",
+ "title": "Jurisdiction"
+ },
+ "legal_name": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's legal name exactly as GLEIF publishes it, verbatim and never reconciled against `CompanyReport.name` or anything else. **Bound by D-028(1)**: never a lookup key, never an index, never searchable, never written to a log (D-040). The binding exists because GLEIF issues LEIs to sole proprietors too — 126,936 records carry an entity-level classification of `SOLE_PROPRIETOR`, whose legal name is routinely a natural person's name — and although no sole-proprietor *parent* was observed in a 30-record sample, nothing in GLEIF's Level 2 format forbids one.",
+ "title": "Legal Name"
+ },
+ "lei": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's own 20-character LEI, when GLEIF discloses a parent for this side. This is a lookup key **for GLEIF**, not for `lookup_company` — the caller cannot feed it a national identifier this project does not carry. To walk up a group from here, call `lookup_company` on the parent's own national identifier where the caller already has it. Norway additionally publishes `CompanyReport.parent_id` on the *first* round trip from Enhetsregisteret — a different register's answer to a neighbouring question, and it is not reconciled against this one (D-018).",
+ "title": "Lei"
+ },
+ "registration_status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The *parent's* own LEI registration status (`ISSUED`, `LAPSED`, …), carrying the same warning as `LeiRecord.registration_status`: a `LAPSED` LEI means the parent stopped renewing its own registration and is **not** evidence of insolvency — see that field's own description rather than restating it here.",
+ "title": "Registration Status"
+ },
+ "reporting_exception": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Why the entity did not report a parent on this side, from GLEIF's closed exception-reason vocabulary, relayed verbatim: `NO_LEI`, `NATURAL_PERSONS` ('the entity is controlled by a natural person(s) without any intermediate legal entity'), `NON_CONSOLIDATING` ('controlled by legal entities not subject to consolidation'), `NO_KNOWN_PERSON` ('no known person(s) controlling the entity, e.g. the entity is controlled by diverse shareholders'), `NON_PUBLIC`, and five values deprecated since 2022-03-01 and retained only for compatibility (`BINDING_LEGAL_COMMITMENTS`, `LEGAL_OBSTACLES`, `DISCLOSURE_DETRIMENTAL`, `DETRIMENT_NOT_EXCLUDED`, `CONSENT_NOT_OBTAINED`). Three things to hold onto reading it, all load-bearing: it is a **category word and never a name** — nothing beyond the word is available and nothing beyond it would be relayed if it were (D-028); it is the entity's **own stated reason**, verified by neither GLEIF nor us; and it is **not applied consistently between filers** — of 117 ultimate-parent exceptions sampled, `NATURAL_PERSONS` is 49 of 57 (86%) of Norwegian ones and 28 of 60 (47%) of British ones, and EQUINOR ASA (`OW6OFBNCKXC4US5C7523`), 67% owned by the Norwegian State, reads `NATURAL_PERSONS` while DNB, Telenor and Tesco — in the same position at the top of their own groups — read `NON_CONSOLIDATING`, and BP, Ericsson and Carillion read `NO_KNOWN_PERSON`. Treat `NATURAL_PERSONS` on a company with a known institutional owner as a filing artefact, not a fact about its owners.",
+ "title": "Reporting Exception"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The exact GLEIF URL that returned this side's own record — the parent's Level 1 record when disclosed, or the reporting-exception record when excepted. `ParentBlock.provenance` describes the fetch as a whole; this names the one leg that filled this side (D-046(d)).",
+ "title": "Source Url"
+ }
+ },
+ "title": "ParentLink",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The entity that directly consolidates this entity's accounts, or the reporting exception recorded for this side. `None` when GLEIF's relationships name neither a disclosed parent nor a reporting exception for this side (0 of 1,800 records sampled) or when a failed fetch degraded just this side (`notes` names the leg); see `ParentBlock`'s own docstring for the two other reasons this can be `None`."
+ },
+ "notes": {
+ "description": "Plain-English caveats, always naming the URL(s) actually fetched — including, when GLEIF holds no LEI at all, a sentence saying so. Whenever either side carries a `reporting_exception`, also carries the 'this word is the filer's own, unverified and inconsistently applied' sentence, because a caveat that lives only in a schema description does not travel into a rendered answer. Also names the one thing lost by not modelling the disclosed relationship's own period history: its start date.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "One `SourceRef` for the whole fan-out (D-046(d)), not one per side: `source` names GLEIF, `license` is 'CC0 1.0', `source_url` is 'https://api.gleif.org/api/v1/lei-records/{lei}' — the record every leg hangs off — and `fetched_at` is the moment the fan-out completed. The discovery search that finds this entity's LEI fills no field on this block; it is shared with `include=[\"lei\"]` and never repeated for a lookup that asks for both. Each side's own `source_url` names the exact leg that filled it. Never implies endorsement and never describes registry-mcp as a GLEIF service (D-026(c)).",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "ultimate": {
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "One side (`direct` or `ultimate`) of a `ParentBlock` — either GLEIF\ndiscloses a corporate parent for this side, or the entity's own filer\nexplained why it did not (``include=[\"parents\"]``, D-047(a)).\n\n**Exactly one of `lei` and `reporting_exception` is populated — never\nboth, never neither** (enforced below, D-011: a link that discloses a\nparent's LEI and a link that records why no parent was disclosed are two\ndifferent facts and must not collapse into one).\n\nGLEIF Level 2 reports the *accounting consolidating* parent — the wire\nword is `IS_DIRECTLY_CONSOLIDATED_BY` / `IS_ULTIMATELY_CONSOLIDATED_BY` —\nwhich is **not** the same as the majority shareholder, and neither\nimplies the other.",
+ "properties": {
+ "corroboration_level": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "GLEIF's own corroboration level for the *relationship* record, verbatim: `FULLY_CORROBORATED`, `PARTIALLY_CORROBORATED`, `ENTITY_SUPPLIED_ONLY`. Measured over 31 disclosed relationships: 10 / 11 / 10 — roughly a third of all disclosed parent links are the filer's own assertion that nobody has checked, which is the reason this field costs a second request. `None` on the exception side, where there is no relationship record.",
+ "title": "Corroboration Level"
+ },
+ "jurisdiction": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's own registered jurisdiction, ISO-3166-1 alpha-2, as GLEIF publishes it. May differ from this entity's own `CompanyReport.country`.",
+ "title": "Jurisdiction"
+ },
+ "legal_name": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's legal name exactly as GLEIF publishes it, verbatim and never reconciled against `CompanyReport.name` or anything else. **Bound by D-028(1)**: never a lookup key, never an index, never searchable, never written to a log (D-040). The binding exists because GLEIF issues LEIs to sole proprietors too — 126,936 records carry an entity-level classification of `SOLE_PROPRIETOR`, whose legal name is routinely a natural person's name — and although no sole-proprietor *parent* was observed in a 30-record sample, nothing in GLEIF's Level 2 format forbids one.",
+ "title": "Legal Name"
+ },
+ "lei": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The parent's own 20-character LEI, when GLEIF discloses a parent for this side. This is a lookup key **for GLEIF**, not for `lookup_company` — the caller cannot feed it a national identifier this project does not carry. To walk up a group from here, call `lookup_company` on the parent's own national identifier where the caller already has it. Norway additionally publishes `CompanyReport.parent_id` on the *first* round trip from Enhetsregisteret — a different register's answer to a neighbouring question, and it is not reconciled against this one (D-018).",
+ "title": "Lei"
+ },
+ "registration_status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The *parent's* own LEI registration status (`ISSUED`, `LAPSED`, …), carrying the same warning as `LeiRecord.registration_status`: a `LAPSED` LEI means the parent stopped renewing its own registration and is **not** evidence of insolvency — see that field's own description rather than restating it here.",
+ "title": "Registration Status"
+ },
+ "reporting_exception": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Why the entity did not report a parent on this side, from GLEIF's closed exception-reason vocabulary, relayed verbatim: `NO_LEI`, `NATURAL_PERSONS` ('the entity is controlled by a natural person(s) without any intermediate legal entity'), `NON_CONSOLIDATING` ('controlled by legal entities not subject to consolidation'), `NO_KNOWN_PERSON` ('no known person(s) controlling the entity, e.g. the entity is controlled by diverse shareholders'), `NON_PUBLIC`, and five values deprecated since 2022-03-01 and retained only for compatibility (`BINDING_LEGAL_COMMITMENTS`, `LEGAL_OBSTACLES`, `DISCLOSURE_DETRIMENTAL`, `DETRIMENT_NOT_EXCLUDED`, `CONSENT_NOT_OBTAINED`). Three things to hold onto reading it, all load-bearing: it is a **category word and never a name** — nothing beyond the word is available and nothing beyond it would be relayed if it were (D-028); it is the entity's **own stated reason**, verified by neither GLEIF nor us; and it is **not applied consistently between filers** — of 117 ultimate-parent exceptions sampled, `NATURAL_PERSONS` is 49 of 57 (86%) of Norwegian ones and 28 of 60 (47%) of British ones, and EQUINOR ASA (`OW6OFBNCKXC4US5C7523`), 67% owned by the Norwegian State, reads `NATURAL_PERSONS` while DNB, Telenor and Tesco — in the same position at the top of their own groups — read `NON_CONSOLIDATING`, and BP, Ericsson and Carillion read `NO_KNOWN_PERSON`. Treat `NATURAL_PERSONS` on a company with a known institutional owner as a filing artefact, not a fact about its owners.",
+ "title": "Reporting Exception"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The exact GLEIF URL that returned this side's own record — the parent's Level 1 record when disclosed, or the reporting-exception record when excepted. `ParentBlock.provenance` describes the fetch as a whole; this names the one leg that filled this side (D-046(d)).",
+ "title": "Source Url"
+ }
+ },
+ "title": "ParentLink",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The entity at the top of the chain that ultimately consolidates this entity's accounts, or the reporting exception recorded for this side. Independent of `direct` — never derived from it, never reconciled against it."
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "ParentBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Corporate parents from GLEIF Level 2 — accounting consolidation, not shareholding (see `ParentBlock`'s own docstring). `None` unless `parents` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for why). A country that declares this attachment returns a *present* block even for an entity GLEIF holds no LEI for at all, with `direct` and `ultimate` both `None`: GLEIF cannot hold Level 2 for an entity it has no Level 1 for, and that is the answer, not an absence (D-011). Declared by default by every country whose identifiers cannot be a natural person's (`Registry.universal_includes`, alongside `lei`) — Sweden does not declare it, for the same reason it does not declare `lei`."
+}Output schema / properties / peppolAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Whether one entity can be reached over the Peppol network — the\n``include=[\"peppol\"]`` attachment (DECISIONS.md D-029(b), amended in full\nby D-046 after ``tasks/T48-recon.md`` read the wire). Norway-only today:\ndeclared by ``BrregRegistry.supported_includes``, not by\n:attr:`Registry.universal_includes` (D-046(h)) — the participant\nidentifier needs a country's own ISO 6523 ICD, the answer's provenance is\na *different SMP per participant* rather than one endpoint, and the\nlicence sentence below was earned by reading a Norwegian catalogue page.\n\nThe Peppol network is not Enhetsregisteret: the answering SMP is operated\nby a second organisation entirely (for Norway, Digitaliseringsdirektoratet\n— named in ``registries/no/peppol.py``, where national vocabulary belongs\nper D-004, never here), so this is a second round trip with its own\n:class:`SourceRef` rather than a field on :class:`CompanyReport` itself\n(D-026(c)).",
+ "properties": {
+ "can_receive_invoice": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this participant advertises the Peppol BIS Billing 3.0 Invoice document type or its PINT successor — DFØ's own equation: 'Peppol BIS billing v3.0 er det samme som EHF-faktura', which is exactly the 1 January 2027 question. Derived by **exact membership of a committed table of document type identifiers**, never by matching text, a substring or a version range. `true` when a table id is in `document_types` via the authoritative SMP route; also `true` via the Peppol Directory fallback, but a Directory list is a subset of the SMP's, so a Directory miss here is `null`, never `false` — see `document_types`. `false` only when `registered` itself is `false`. `null` when `registered` is `null`. **The SMP publishes a per-document-type ServiceActivationDate/ServiceExpirationDate that this block does not read** (one extra HTTP call per document type), so an advertised document type may be future-dated or already expired — `notes` says so whenever this is `true`.",
+ "title": "Can Receive Invoice"
+ },
+ "document_types": {
+ "description": "The **document type** identifiers the answering SMP (or, on the Directory fallback, the Peppol Directory) advertises for this participant, full qualified `'<scheme>::<value>'` strings, e.g. `'busdox-docid-qns::urn:oasis:...:billing:3.0::2.1'`. **Not process identifiers** — those live one HTTP call deeper, per document type, and are not carried (D-046(e)). No cap and no truncation: the modal Norwegian participant lists two, some list many more. Only *receiving* capabilities are registered anywhere in the Peppol network, which is the right semantics for 'can this counterparty receive an e-invoice'.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Document Types",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block. Always names which route answered (the SMP, or the Peppol Directory) or which step failed when `registered` is `null`; that only *receiving* capabilities are registered in the Peppol network; on the Directory route, that it is a voluntary, lagging index of the SMP; and, on a Directory miss, the operators' own statement that this does not mean the entity is not in the Peppol Network.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "participant_id": {
+ "description": "The ISO 6523 participant identifier, `'0192:' + normalised orgnr` — `0192` is Norway's ICD (International Code Designator) inside the Peppol network. Derived offline from the identifier alone and **always populated, even when every lookup failed** (D-029(c)): it is the key a caller needs to ask elsewhere, regardless of what this block's own `registered` field says.",
+ "title": "Participant Id",
+ "type": "string"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block's answer was produced. `source` names the SMP host and route that answered (e.g. '<host> (Peppol SMP, via the Peppol SML)'), or the Peppol Directory named as an index that may lag — **derived at request time from what actually answered, never a constant** (D-046(b)). `license` carries D-046(g)'s stated absence: nobody publishes a licence for the Peppol SML, for any SMP or for the Peppol Directory. One `SourceRef` for the whole block even though up to two round trips were made (a DNS read and an HTTPS read): the DNS step only located the host and fills no field of this block except `smp_url`, so it is disclosed as that field rather than as a second provenance (D-046(d)).",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data for this attachment's own fetch, e.g. 'CC0 1.0'. A block's licence is the register's own for that endpoint and may equal the base report's without being derived from it: Norway's Regnskapsregisteret accounts endpoint states no licence of its own, so the blocks it serves carry NLOD 2.0 — the same value the company record on the same host already carries.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "registered": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Three states, and an agent branches on this field under a statute, so all three matter. `true`: the Peppol network answered for this participant — either the SMP the Peppol SML named for it served a ServiceGroup, or the Peppol Directory listed a match. `false`: **the authoritative SML/SMP route answered that it is not registered** — an NXDOMAIN resolving the Peppol SML, or a 404 from the SMP the SML named. Nothing else ever earns `false` (D-046(a)). `null`: we could not get an authoritative answer — a DNS resolver exception, a timeout, a NOERROR answer with no `Meta:SMP` record, an SMP error, **or the Peppol Directory simply not listing this participant**, which both Peppol operators state in writing means nothing: publication to the Directory is voluntary, and its own introduction page says a miss there 'doesn't mean the entity is not in the Peppol Network' (D-046(a)).",
+ "title": "Registered"
+ },
+ "smp_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The SMP base URL the Peppol SML named for **this participant**, verbatim. This varies per participant and is **never assumed**: the default national SMP a country's participants mostly resolve to is not the only one — a measured 1-in-43 Norwegian participants resolve to a different one entirely (D-046(b)). `null` when the SML never named a host for this participant (NXDOMAIN, or no usable `Meta:SMP` record).",
+ "title": "Smp Url"
+ }
+ },
+ "required": [
+ "participant_id",
+ "provenance"
+ ],
+ "title": "PeppolParticipant",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this entity can be reached over the Peppol e-invoicing network. `None` unless `peppol` was passed in `include=[...]` — and, even then, `None` only if the attachment could not be built at all (see `notes` for why). A country that declares this attachment returns a **present** block even when the network could not be reached: `PeppolParticipant.registered` carries the three-state answer (`true`/`false`/`null`) and `PeppolParticipant.participant_id` is always populated, because that is the key a caller needs to ask elsewhere regardless (D-011, D-029(c)). Norway only, today (D-046(h)): the participant identifier needs a country's own ISO 6523 ICD and the answer's provenance is a different SMP per participant, neither of which generalises to `Registry.universal_includes` yet."
+}search_company2 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."Input schema / properties / country / examplesPrevious value: -[
- "NO",
- "GB"
-]New value: +[
+ "NO",
+ "GB",
+ "SE"
+]validate_company_id3 fields changedInput schema / properties / country / descriptionPrevious value: -"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."New value: +"ISO-3166-1 alpha-2 country code. 'NO' = Norway (Brønnøysundregistrene / Enhetsregisteret), 'GB' = United Kingdom (Companies House), 'SE' = Sweden (Bolagsverket). 'UK' is not a country code here and is rejected. Call list_countries for the current set rather than hard-coding one."Input schema / properties / country / examplesPrevious value: -[
- "NO",
- "GB"
-]New value: +[
+ "NO",
+ "GB",
+ "SE"
+]Input schema / properties / id / descriptionPrevious value: -"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you."New value: +"The company's national identifier. Norway (country='NO'): a nine-digit organisasjonsnummer (orgnr), e.g. '923609016'; spaces, dots and a 'NO...MVA' VAT suffix are accepted and normalised. United Kingdom (country='GB'): a Companies House company number (CRN), eight characters, e.g. '00445790' or 'OC303675'; a short number is zero-padded for you. Sweden (country='SE'): a ten-digit organisationsnummer, e.g. '5560160680', or a twelve-digit personnummer for a sole trader."lookup_company5 fields changedInput schema / properties / include / descriptionPrevious value: -"Optional attachment names to fetch alongside the base report, e.g. ['charges'] for United Kingdom (country='GB') registered charges (mortgages and other security interests) — a second, independent fetch attached at that same name, with its own provenance. Empty by default, which costs exactly one upstream request, same as before this argument existed. Most countries declare none yet — call list_countries and check a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result."New value: +"Optional attachment names to fetch alongside the base report. Each is a second, independent fetch attached at that same name on the result, with its own provenance, and null unless you ask for it. Three exist today: 'filings' (what the entity has actually filed, and when — every country), 'charges' (mortgages and other security interests, United Kingdom only) and 'insolvency' (winding-up and administration proceedings, United Kingdom only). Empty by default, which costs exactly one upstream request. Ask for one when the base report is not enough to answer the question in front of you: 'filings' answers whether they file on time, 'charges' whether assets are already pledged, 'insolvency' whether they are being wound up. Call list_countries and read a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result."Input schema / properties / include / examplesPrevious value: -[
- [
- "charges"
- ],
- []
-]New value: +[
+ [
+ "filings"
+ ],
+ [
+ "charges",
+ "insolvency"
+ ],
+ []
+]Output schema / properties / charges / anyOfPrevious value: -[
- {
- "additionalProperties": false,
- "description": "Registered charges for one entity — an `include=[\"charges\"]` attachment\n(D-042(g),(h)), never a plain field on `CompanyReport` (D-041(c)): it is a\nsecond round trip with its own moment, its own cache state and its own\nfailure mode, so it carries its own `SourceRef` rather than reusing the\nreport's.\n\nTwo-level nullability is the point of this shape (D-026(c), D-041(c),\nD-042(d)(3)): `CompanyReport.charges` is `None` when `charges` was not in\n`include`, or when the fetch failed (`Registry.lookup_with` appends a\n`notes` sentence on the report saying which). Once *present*, this block\ncarries `charges: []` for an entity the register confirms has none —\nthat case must never collapse into the absent case, and must never be\n`not_found` (D-011).",
- "properties": {
- "charges": {
- "description": "Sorted newest first (a country module's own tie-break rule).",
- "items": {
- "additionalProperties": false,
- "description": "One registered charge (a mortgage or other security interest) against\nan entity — one row of a `ChargeBlock`.\n\nField names are country-neutral (D-042(g)): GB is the first filler\n(Companies House `/company/{n}/charges`, `registries/gb/__init__.py`),\nand any future filler (e.g. Norway's Løsøreregisteret, once it opens a\npublic API) maps onto this same shape rather than getting one of its own.",
- "properties": {
- "assets_charged": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own free-text description of what is charged.",
- "title": "Assets Charged"
- },
- "charge_id": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own opaque handle for this charge; not fetchable through this API. `None` when the register has no such handle for an older filing — honestly absent, not guessed.",
- "title": "Charge Id"
- },
- "charge_number": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's sequence number for this charge, within this company.",
- "title": "Charge Number"
- },
- "classification": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "What kind of instrument this is, as the register describes it.",
- "title": "Classification"
- },
- "contains_floating_charge": {
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Whether the register marks this instrument as including a floating charge.",
- "title": "Contains Floating Charge"
- },
- "created_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was created.",
- "title": "Created On"
- },
- "delivered_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was delivered to the register for registration.",
- "title": "Delivered On"
- },
- "is_outstanding": {
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Derived from `status` by membership of a country module's own committed table of status words it has actually observed on the wire. `None` when `status` is absent or is a word not yet in that table — never guessed, never `False` by default (D-025(d), D-011).",
- "title": "Is Outstanding"
- },
- "obligations_secured": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own free-text description of what the charge secures.",
- "title": "Obligations Secured"
- },
- "parties_entitled": {
- "description": "Names exactly as the register publishes them for the party or parties the charge is entitled to (typically a bank, an insurer or a trustee company; occasionally a natural person, e.g. a director lending to their own company). This is a term of the company's own instrument, not a person record: it is never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040). It is the one place in this product a natural person's name can appear, and it is deliberately not named the register's own `persons_entitled` — that name asserts a natural person; this one does not.",
- "items": {
- "type": "string"
- },
- "title": "Parties Entitled",
- "type": "array"
- },
- "satisfied_on": {
- "anyOf": [
- {
- "format": "date",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "When the charge was satisfied, if it has been.",
- "title": "Satisfied On"
- },
- "status": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own word, verbatim, e.g. \"outstanding\", \"fully-satisfied\" — national vocabulary lives here, in the value, never in a field name (D-042(g)). See `is_outstanding` for the country-neutral derived flag.",
- "title": "Status"
- }
- },
- "title": "Charge",
- "type": "object"
- },
- "title": "Charges",
- "type": "array"
- },
- "notes": {
- "description": "Plain-English caveats about this block, e.g. truncation when `total_count` exceeds `len(charges)`.",
- "items": {
- "type": "string"
- },
- "title": "Notes",
- "type": "array"
- },
- "outstanding_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Derived as `total_count - satisfied_count` when the register publishes both as whole-company figures; a register that also tracks a distinct partially-satisfied state folds it in here as not-fully-satisfied. `None` when the register does not publish enough to derive it.",
- "title": "Outstanding Count"
- },
- "provenance": {
- "additionalProperties": false,
- "description": "Where, when and under what licence this block was fetched.",
- "properties": {
- "cached": {
- "default": false,
- "description": "True when this attachment was served from cache rather than a live fetch.",
- "title": "Cached",
- "type": "boolean"
- },
- "fetched_at": {
- "anyOf": [
- {
- "format": "date-time",
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "UTC timestamp of the live fetch this attachment came from.",
- "title": "Fetched At"
- },
- "license": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
- "title": "License"
- },
- "source": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
- "title": "Source"
- },
- "source_url": {
- "anyOf": [
- {
- "type": "string"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Direct URL of the upstream record, for citation.",
- "title": "Source Url"
- }
- },
- "title": "SourceRef",
- "type": "object"
- },
- "satisfied_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own whole-company count of satisfied charges, verbatim.",
- "title": "Satisfied Count"
- },
- "total_count": {
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "The register's own count of charges for this company, which may exceed `len(charges)` — see `notes` for a truncation disclosure when it does.",
- "title": "Total Count"
- }
- },
- "required": [
- "provenance"
- ],
- "title": "ChargeBlock",
- "type": "object"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "additionalProperties": false,
+ "description": "Registered charges for one entity — an `include=[\"charges\"]` attachment\n(D-042(g), D-045(a)), never a plain field on `CompanyReport` (D-041(c)):\nit is a second round trip with its own moment, its own cache state and its\nown failure mode, so it carries its own `SourceRef` rather than reusing\nthe report's.\n\n**The count fields are ruled by D-045(a)**, not by D-042(h). It strikes\n`outstanding_count` — `total_count - satisfied_count` is our arithmetic\nwearing a register figure's name, and on a shared model it would mean\n\"outstanding\" for a register with no partially-satisfied state and \"not\nfully satisfied\" for one that has it (D-011) — and adds the register's\nown `part_satisfied_count` in its place. **This class has not yet been\nreconciled with that ruling**; the task that does so is named in D-045(a)\nand is due before the next deploy.\n\nTwo-level nullability is the point of this shape (D-026(c), D-041(c),\nD-042(d)(3)): `CompanyReport.charges` is `None` when `charges` was not in\n`include`, or when the fetch failed (`Registry.lookup_with` appends a\n`notes` sentence on the report saying which). Once *present*, this block\ncarries `charges: []` for an entity the register confirms has none —\nthat case must never collapse into the absent case, and must never be\n`not_found` (D-011).",
+ "properties": {
+ "charges": {
+ "description": "Sorted newest first (a country module's own tie-break rule).",
+ "items": {
+ "additionalProperties": false,
+ "description": "One registered charge (a mortgage or other security interest) against\nan entity — one row of a `ChargeBlock`.\n\nField names are country-neutral (D-042(g)): GB is the first filler\n(Companies House `/company/{n}/charges`, `registries/gb/__init__.py`),\nand any future filler (e.g. Norway's Løsøreregisteret, once it opens a\npublic API) maps onto this same shape rather than getting one of its own.\n\n**The field list is ruled by D-045(a)**, not by D-042(h) — which rules\n`FiledDocument` and no charge shape at all. D-045(a) accepts these names,\ncorrects `created_on`'s description (it is the date the charge instrument\nwas created, not a record timestamp), and adds `contains_fixed_charge` and\n`contains_negative_pledge` beside `contains_floating_charge`, all three of\nwhich Companies House emits **only when true** — so an absent flag means\nthe register did not mark this instrument, never that it lacks one\n(D-011). **This class has not yet been reconciled with that ruling**; the\ntask that does so is named in D-045(a) and is due before the next deploy.",
+ "properties": {
+ "assets_charged": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what is charged.",
+ "title": "Assets Charged"
+ },
+ "charge_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this charge; not fetchable through this API. `None` when the register has no such handle for an older filing — honestly absent, not guessed.",
+ "title": "Charge Id"
+ },
+ "charge_number": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's sequence number for this charge, within this company.",
+ "title": "Charge Number"
+ },
+ "classification": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What kind of instrument this is, as the register describes it.",
+ "title": "Classification"
+ },
+ "contains_floating_charge": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether the register marks this instrument as including a floating charge.",
+ "title": "Contains Floating Charge"
+ },
+ "created_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was created.",
+ "title": "Created On"
+ },
+ "delivered_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was delivered to the register for registration.",
+ "title": "Delivered On"
+ },
+ "is_outstanding": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Derived from `status` by membership of a country module's own committed table of status words it has actually observed on the wire. `None` when `status` is absent or is a word not yet in that table — never guessed, never `False` by default (D-025(d), D-011).",
+ "title": "Is Outstanding"
+ },
+ "obligations_secured": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what the charge secures.",
+ "title": "Obligations Secured"
+ },
+ "parties_entitled": {
+ "description": "Names exactly as the register publishes them for the party or parties the charge is entitled to (typically a bank, an insurer or a trustee company; occasionally a natural person, e.g. a director lending to their own company). This is a term of the company's own instrument, not a person record: it is never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040). It is the one place in this product a natural person's name can appear, and it is deliberately not named the register's own `persons_entitled` — that name asserts a natural person; this one does not.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Parties Entitled",
+ "type": "array"
+ },
+ "satisfied_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was satisfied, if it has been.",
+ "title": "Satisfied On"
+ },
+ "status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word, verbatim, e.g. \"outstanding\", \"fully-satisfied\" — national vocabulary lives here, in the value, never in a field name (D-042(g)). See `is_outstanding` for the country-neutral derived flag.",
+ "title": "Status"
+ }
+ },
+ "title": "Charge",
+ "type": "object"
+ },
+ "title": "Charges",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block, e.g. truncation when `total_count` exceeds `len(charges)`.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "outstanding_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Derived as `total_count - satisfied_count` when the register publishes both as whole-company figures; a register that also tracks a distinct partially-satisfied state folds it in here as not-fully-satisfied. `None` when the register does not publish enough to derive it.",
+ "title": "Outstanding Count"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "satisfied_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own whole-company count of satisfied charges, verbatim.",
+ "title": "Satisfied Count"
+ },
+ "total_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own count of charges for this company, which may exceed `len(charges)` — see `notes` for a truncation disclosure when it does.",
+ "title": "Total Count"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "ChargeBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+]Output schema / properties / filingsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "What one entity has filed with its national register — an\n``include=[\"filings\"]`` attachment (D-041(d), D-042), never a plain field\non :class:`CompanyReport` (D-041(c)): it is a second round trip with its\nown moment, its own cache state and its own failure mode, so it carries its\nown :class:`SourceRef` rather than reusing the report's.\n\nAll three live countries declare it, and each answers a differently-scoped\nquestion its register actually supports — Companies House returns the whole\nfiling history, Bolagsverket the filed annual reports, Regnskapsregisteret\nthe filed annual accounts. `notes` says which, in words, on every block.\n\nTwo-level nullability is the contract (D-011, D-026(c), D-041(c)):\n**absent** means \"you did not ask, or the fetch failed\" — `lookup_with`\nappends one `notes` sentence to the report saying which — while **present\nwith `documents: []`** means \"the register lists no filings for this\nentity\", a real and useful answer about a counterparty that must never be\nrendered as an absence.",
+ "properties": {
+ "documents": {
+ "description": "One page of the register's own filing history, newest first by `filed_at`. Never paginated further; when the register holds more, `total_count` says how many and `notes` says so in words. Empty means the register lists none — not that we could not look.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One filing a national register publishes for one entity — one row of a\n:class:`FilingHistory`.\n\nThe shape ``DECISIONS.md`` D-041(d) ruled and D-042(h) widened, and it is\ncountry-neutral by construction rather than by intent: Britain, Sweden and\nNorway each built this model independently behind their own seam, and all\nthree arrived field-for-field at this one. National vocabulary lives in the\n*values* (`category`, `type_code`, `description_code`), never in a field\nname (D-042(g)).\n\nEvery field is nullable and every `None` means the same thing: **the\nregister does not publish it** (D-011). It never means zero, never means\n\"no\", and is never filled by derivation — a register that does not publish\na period start gets `None`, not a start inferred by subtracting twelve\nmonths from the end (D-009).",
+ "properties": {
+ "category": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own category for this filing, verbatim and never translated — \"accounts\", \"mortgage\", \"confirmation-statement\", \"gazette\" and some twenty more in Britain alone, and none of these lists is closed. It is the field `kind` is derived from.",
+ "title": "Category"
+ },
+ "days_from_fee_point": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Signed days from a named late-fee datum to `filed_at`, **only where the register itself publishes such a datum for that period**. Negative is early. `None` is the common answer and means the datum does not exist in the data, not that the arithmetic was skipped: Sweden fills it because årsredovisningslagen 8 kap. 6 § names one datum for every company, while Companies House publishes only the *next* period's due date and nothing per-period historical, so there is nothing to measure a past filing against. Deriving one from the statutory rule would require guessing a period length, a first-accounts variant and any shortening the register has not disclosed, then presenting the result as the register's own — the invented figure D-009 forbids.",
+ "title": "Days From Fee Point"
+ },
+ "description_code": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own description-template key, verbatim and **never resolved into prose**. This is the key and not the sentence on purpose, and the reason is the whole design of this block: Companies House resolves these templates from a `description_values` object, 97 templates interpolate an officer's name and 26 a person with significant control's, so the resolved sentence is personal data while the key is not. **The key says what happened; only the values say who** (D-042(e)(1), D-028).",
+ "title": "Description Code"
+ },
+ "document_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this filing, relayed verbatim and never interpreted. **Not fetchable through this API**: the filed document itself lives behind a separate host, which is a second upstream with its own provenance and out of scope for this block (D-041(c)). It is the key a support case with the register can name.",
+ "title": "Document Id"
+ },
+ "file_format": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What the register holds the document as, where it says. `None` where the filing-history endpoint publishes no format — Companies House keeps the media type on its separate document host, a second fetch this block does not make.",
+ "title": "File Format"
+ },
+ "filed_at": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the register recorded this filing, verbatim. This is the field that makes the block answer *does this company file on time*, and it is the sort key for `FilingHistory.documents`: newest first.",
+ "title": "Filed At"
+ },
+ "kind": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The `Deadline.kind` slug this filing discharges, or `None` when it discharges none. This is the one field that is *derived* rather than relayed, and it is derived only by a committed per-country table of category words actually observed on the wire. A filing whose category is outside that table gets `None` rather than an invented slug (D-009): a filing that discharges no deadline this product publishes says so honestly.",
+ "title": "Kind"
+ },
+ "period_end": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's last day, exactly as the register published it. Beware what the period belongs to: on an annual-accounts filing it is the date the accounts were made up to, but a register may publish a made-up date on other filing kinds too — a British confirmation statement carries one, and it is not a financial year end. Read it together with `kind`. `None` on the great majority of filings, which have no reporting period at all.",
+ "title": "Period End"
+ },
+ "period_start": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The reporting period's first day, as published. `None` wherever the register publishes no counterpart to `period_end` — deriving one would assert a period length the register never stated, and a first, shortened or extended accounting period is lawful and common (D-009). Norway's Regnskapsregisteret publishes `regnskapsperiode: {fraDato, tilDato}` and fills both ends; Companies House publishes only the end.",
+ "title": "Period Start"
+ },
+ "type_code": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own form code for this filing, verbatim: \"AA\", \"CS01\", \"AP01\", \"MR01\" and older forms such as \"288a\" and \"363s\" in Britain — 100 distinct codes across 1876 items observed live.",
+ "title": "Type Code"
+ }
+ },
+ "title": "FiledDocument",
+ "type": "object"
+ },
+ "title": "Documents",
+ "type": "array"
+ },
+ "financial_year_end": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The latest reporting period among this entity's filed **annual accounts** (`kind == \"annual_accounts\"`), carried verbatim — never a synthesised month-day, and never taken from a filing of another kind that happens to carry a made-up date of its own. It is the latest *period*, not the period of the latest *filing*, because a register may accept a later filing that amends an earlier year and that would otherwise roll this date backwards. It is **evidence of** the entity's accounting reference date, not a statement of it. `None` when this page holds no annual-accounts filing with a reporting period, including when older accounts exist further back than the page reaches.",
+ "title": "Financial Year End"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block: which subset of filings this register publishes, truncation when `total_count` exceeds `len(documents)`, and which empty state an empty `documents` is.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "total_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own count of filings for this entity, which may greatly exceed `len(documents)` — 8371 against a 25-row page, for one company observed live. **`None` means the register published no count**, not zero, and for Companies House it additionally distinguishes a real zero from a number whose filing history the register cannot serve at all: that endpoint returns `0` for both, and relaying the second as a zero would assert something the register never said (D-011). `notes` names which case it was.",
+ "title": "Total Count"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "FilingHistory",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What this entity has filed with its register, and when. `None` unless `filings` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). Scope differs by country because each register publishes a different subset, and the block's own `notes` says which: Companies House the whole filing history, Bolagsverket the filed annual reports, Regnskapsregisteret the filed annual accounts. A *present* block with `documents: []` means the register lists none — never the same as absent (D-011, D-042(d))."
+}Output schema / properties / insolvencyAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Insolvency proceedings a register publishes against one entity — an\n``include=[\"insolvency\"]`` attachment (D-042), never a plain field on\n:class:`CompanyReport` (D-041(c)), and carrying its own :class:`SourceRef`.\n\nTwo-level nullability is the point of the shape (D-011, D-026(c),\nD-041(c), D-042(d)(3)): once *present*, this block carries ``cases: []``\nfor an entity the register publishes no insolvency case for — that state\nmust never collapse into the absent state and must never be ``not_found``.\nCompanies House's 404 here is the *normal* answer for a solvent company and\nis byte-identical to its answer for a number that was never issued, so it\nsays nothing about whether the entity exists; ``notes`` therefore\ndistinguishes \"the register holds no insolvency resource here\" from \"the\nresource exists and is empty\" instead of flattening both into silence.",
+ "properties": {
+ "cases": {
+ "description": "Every insolvency case the register publishes for this entity — the whole history, not a page, where the register's endpoint is unpaginated. Sorted newest first by the case's most recent event date, then by `case_number` descending; cases the register gives no date for sort last.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One insolvency case a register publishes against one entity.\n\n**No practitioner particular can land here.** A register commonly publishes\neach appointed practitioner's name and postal address alongside the case;\nD-042(e)(2) bars relaying them in the first tranche, so this model has no\nfield for them and no country mapper reads the key. Adding them later is a\ndecision with its own entry in ``DECISIONS.md``, inheriting D-028's four\npreconditions in full.",
+ "properties": {
+ "case_number": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own identifier for this case, verbatim. For Companies House this is a per-company sequence number rendered as a string (\"1\", \"2\", … up to \"31\" in the live sample) and is **not** a court reference — it identifies the case only within this entity. Kept as a string because another register's case identifier need not be numeric.",
+ "title": "Case Number"
+ },
+ "case_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word for the kind of procedure, verbatim — national vocabulary in the value (D-042(g)). Ten words observed live in Britain, among them \"compulsory-liquidation\", \"creditors-voluntary-liquidation\", \"members-voluntary-liquidation\" and \"in-administration\". See `is_liquidation` for the country-neutral derived flag.",
+ "title": "Case Type"
+ },
+ "events": {
+ "description": "The register's own dated steps in this case, newest first. Frequently empty — 172 of the 1,485 live British cases carried no date at all, most of them old receiverships — and an empty list means the register publishes no date for this case, never that nothing happened.",
+ "items": {
+ "additionalProperties": false,
+ "description": "One dated step in an insolvency case, as the register itself records it.\n\nThese are the register's own events, not this service's interpretation of\nthem: D-042(e)(2) rules that case type, case number and *these* dated\nevents carry the entire distress signal a pre-contract check needs.",
+ "properties": {
+ "event_type": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word for what happened, verbatim — national vocabulary lives here, in the value, never in a field name (D-042(g)). Thirteen words have been observed live in Britain, from \"petitioned-on\" and \"wound-up-on\" to \"declaration-solvent-on\" and \"dissolved-on\". A word outside the observed set is still relayed verbatim: this field is never filtered, only reported.",
+ "title": "Event Type"
+ },
+ "occurred_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The date the register gives for this event.",
+ "title": "Occurred On"
+ }
+ },
+ "title": "InsolvencyEvent",
+ "type": "object"
+ },
+ "title": "Events",
+ "type": "array"
+ },
+ "is_liquidation": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether this procedure is a winding-up — the country-neutral question behind the national word in `case_type`. Derived by membership of a committed table of words the country module has actually observed on the wire. `None` when `case_type` is absent or is a word not yet in that table — never guessed, never `False` by default (D-011, D-025(d)). **`True` does not mean insolvent**: a members' voluntary liquidation is a *solvent* winding-up, begun by a declaration of solvency, and 56 of the 1,485 live British cases behind this table were exactly that. Read it as 'the entity is being wound up', not as 'the entity cannot pay'.",
+ "title": "Is Liquidation"
+ },
+ "note_codes": {
+ "description": "The register's own note **codes** for this case, verbatim and never resolved into prose — the same treatment D-042(e)(1) gives a filing's `description_code`. Companies House declares this field an unbounded `array[string]`, so it is the one place in that payload a name could hide; codes are therefore relayed through an allow-list of observed codes, and an unrecognised one is dropped and disclosed in the block's `notes` rather than passed through.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Note Codes",
+ "type": "array"
+ }
+ },
+ "title": "InsolvencyCase",
+ "type": "object"
+ },
+ "title": "Cases",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block: which of the register's two empty states this is, that practitioner particulars exist upstream and are deliberately not relayed, and any note code withheld by the allow-list.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "statuses": {
+ "description": "The register's own entity-level insolvency status words, verbatim — national vocabulary in values (D-042(g)). Eight observed live in Britain, among them \"in-administration\", \"liquidation\" and \"voluntary-arrangement\". An **empty list means the register publishes no such word for this entity**, which is not the same as 'not currently insolvent': about one in ten companies whose Companies House status is itself an insolvency status still has no word here. No yes/no flag is derived from this field for exactly that reason (D-011).",
+ "items": {
+ "type": "string"
+ },
+ "title": "Statuses",
+ "type": "array"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "InsolvencyBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Insolvency proceedings the register publishes against this entity. `None` unless `insolvency` was passed in `include=[...]`, or that fetch failed. A *present* block with `cases: []` means the register publishes no case, which for Companies House is the normal answer for a solvent company — and is not evidence the entity exists, since that register answers the same way for a number never issued. Read `InsolvencyCase.is_liquidation` with its own caveat: a members' voluntary liquidation is a solvent wind-up."
+}lookup_company2 fields changedInput schema / properties / includeAdded value: +{
+ "default": [],
+ "description": "Optional attachment names to fetch alongside the base report, e.g. ['charges'] for United Kingdom (country='GB') registered charges (mortgages and other security interests) — a second, independent fetch attached at that same name, with its own provenance. Empty by default, which costs exactly one upstream request, same as before this argument existed. Most countries declare none yet — call list_countries and check a country's supported_includes before guessing; an include value that country does not declare is a bad_request naming what it does support, never a silently empty result.",
+ "examples": [
+ [
+ "charges"
+ ],
+ []
+ ],
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+}Output schema / properties / chargesAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": false,
+ "description": "Registered charges for one entity — an `include=[\"charges\"]` attachment\n(D-042(g),(h)), never a plain field on `CompanyReport` (D-041(c)): it is a\nsecond round trip with its own moment, its own cache state and its own\nfailure mode, so it carries its own `SourceRef` rather than reusing the\nreport's.\n\nTwo-level nullability is the point of this shape (D-026(c), D-041(c),\nD-042(d)(3)): `CompanyReport.charges` is `None` when `charges` was not in\n`include`, or when the fetch failed (`Registry.lookup_with` appends a\n`notes` sentence on the report saying which). Once *present*, this block\ncarries `charges: []` for an entity the register confirms has none —\nthat case must never collapse into the absent case, and must never be\n`not_found` (D-011).",
+ "properties": {
+ "charges": {
+ "description": "Sorted newest first (a country module's own tie-break rule).",
+ "items": {
+ "additionalProperties": false,
+ "description": "One registered charge (a mortgage or other security interest) against\nan entity — one row of a `ChargeBlock`.\n\nField names are country-neutral (D-042(g)): GB is the first filler\n(Companies House `/company/{n}/charges`, `registries/gb/__init__.py`),\nand any future filler (e.g. Norway's Løsøreregisteret, once it opens a\npublic API) maps onto this same shape rather than getting one of its own.",
+ "properties": {
+ "assets_charged": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what is charged.",
+ "title": "Assets Charged"
+ },
+ "charge_id": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own opaque handle for this charge; not fetchable through this API. `None` when the register has no such handle for an older filing — honestly absent, not guessed.",
+ "title": "Charge Id"
+ },
+ "charge_number": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's sequence number for this charge, within this company.",
+ "title": "Charge Number"
+ },
+ "classification": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "What kind of instrument this is, as the register describes it.",
+ "title": "Classification"
+ },
+ "contains_floating_charge": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether the register marks this instrument as including a floating charge.",
+ "title": "Contains Floating Charge"
+ },
+ "created_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was created.",
+ "title": "Created On"
+ },
+ "delivered_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was delivered to the register for registration.",
+ "title": "Delivered On"
+ },
+ "is_outstanding": {
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Derived from `status` by membership of a country module's own committed table of status words it has actually observed on the wire. `None` when `status` is absent or is a word not yet in that table — never guessed, never `False` by default (D-025(d), D-011).",
+ "title": "Is Outstanding"
+ },
+ "obligations_secured": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own free-text description of what the charge secures.",
+ "title": "Obligations Secured"
+ },
+ "parties_entitled": {
+ "description": "Names exactly as the register publishes them for the party or parties the charge is entitled to (typically a bank, an insurer or a trustee company; occasionally a natural person, e.g. a director lending to their own company). This is a term of the company's own instrument, not a person record: it is never a lookup key, never indexed, never searchable and never reaches a log line (D-028(1), D-040). It is the one place in this product a natural person's name can appear, and it is deliberately not named the register's own `persons_entitled` — that name asserts a natural person; this one does not.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Parties Entitled",
+ "type": "array"
+ },
+ "satisfied_on": {
+ "anyOf": [
+ {
+ "format": "date",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "When the charge was satisfied, if it has been.",
+ "title": "Satisfied On"
+ },
+ "status": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own word, verbatim, e.g. \"outstanding\", \"fully-satisfied\" — national vocabulary lives here, in the value, never in a field name (D-042(g)). See `is_outstanding` for the country-neutral derived flag.",
+ "title": "Status"
+ }
+ },
+ "title": "Charge",
+ "type": "object"
+ },
+ "title": "Charges",
+ "type": "array"
+ },
+ "notes": {
+ "description": "Plain-English caveats about this block, e.g. truncation when `total_count` exceeds `len(charges)`.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Notes",
+ "type": "array"
+ },
+ "outstanding_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Derived as `total_count - satisfied_count` when the register publishes both as whole-company figures; a register that also tracks a distinct partially-satisfied state folds it in here as not-fully-satisfied. `None` when the register does not publish enough to derive it.",
+ "title": "Outstanding Count"
+ },
+ "provenance": {
+ "additionalProperties": false,
+ "description": "Where, when and under what licence this block was fetched.",
+ "properties": {
+ "cached": {
+ "default": false,
+ "description": "True when this attachment was served from cache rather than a live fetch.",
+ "title": "Cached",
+ "type": "boolean"
+ },
+ "fetched_at": {
+ "anyOf": [
+ {
+ "format": "date-time",
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "UTC timestamp of the live fetch this attachment came from.",
+ "title": "Fetched At"
+ },
+ "license": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Licence of the upstream data, e.g. 'CC0 1.0'.",
+ "title": "License"
+ },
+ "source": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Human-readable source name, e.g. 'GLEIF Level 1 (gleif.org)'.",
+ "title": "Source"
+ },
+ "source_url": {
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Direct URL of the upstream record, for citation.",
+ "title": "Source Url"
+ }
+ },
+ "title": "SourceRef",
+ "type": "object"
+ },
+ "satisfied_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own whole-company count of satisfied charges, verbatim.",
+ "title": "Satisfied Count"
+ },
+ "total_count": {
+ "anyOf": [
+ {
+ "type": "integer"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "The register's own count of charges for this company, which may exceed `len(charges)` — see `notes` for a truncation disclosure when it does.",
+ "title": "Total Count"
+ }
+ },
+ "required": [
+ "provenance"
+ ],
+ "title": "ChargeBlock",
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Registered charges (mortgages / security interests) against this entity. `None` unless `charges` was passed in `include=[...]` — and, even then, `None` if that fetch failed (see `notes` for which attachment and why). A country that declares this attachment (`CountryInfo.supported_includes`) returns a *present* block with an empty `charges` list for an entity that genuinely has none — the two states never collapse into each other (D-011, D-042(d))."
+}list_countries1 field changedOutput schema / properties / countries / items / properties / supported_includesAdded value: +{
+ "description": "Attachment names this registry declares — the closed set of valid values for `include=[…]` on `lookup_company`, sorted. Empty when this registry offers no attachments today. Lets an agent discover what more it can ask for before it asks (DECISIONS.md D-042(d)); an `include` value outside this list raises `bad_request` naming this same set, never a silently empty block.",
+ "items": {
+ "type": "string"
+ },
+ "title": "Supported Includes",
+ "type": "array"
+}lookup_company2 fields changedOutput schema / properties / advertising_protectedAdded value: +{
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Whether the register marks this entity as protected against direct-marketing use (Danish CVR-loven § 19 'reklamebeskyttelse', Swedish 'reklamspärr'). True: the register marks it. False: the register publishes such a flag for this entity and it is not set. None: this register publishes no such flag at all — the default, and it must never default to False, since False asserts a claim about a register that made none. When True, a country module must also append a `notes` entry containing the phrase 'direct marketing' (case-insensitive) stating the protection — that phrase is the contract this model enforces (see the validator below) — because the marking is a legal condition of passing this record's contact details on, and it must travel with them.",
+ "title": "Advertising Protected"
+}Output schema / properties / euidAdded value: +{
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "European Unique Identifier (EUID, Commission Implementing Regulation (EU) 2021/1042 Article 9), where the register publishes one, e.g. Finland's 'FIFPRO.0112038-9'. None for a register that does not (today: all of ours). Three traps: (1) this is not the LEI — the EUID is register-issued, mandatory in the EU and free, the LEI is voluntary, global, LOU-issued and fee-bearing; an entity may carry both, one or neither. (2) 'EUid' also names the EU Digital Identity wallet, a personal credential unrelated to company registers. (3) it is not stable across a register reorganisation, since it encodes the register of origin (e.g. France's RNE replacing the RCS in 2023). Carried verbatim from the register; never constructed from parts.",
+ "title": "Euid"
+}company_deadlinesfetchlist_countrieslookup_companysearchsearch_companyvalidate_company_idBrønnøysund Register Centre (BRREG) MCP — Norway's official business register.
Finnish Patent and Registration Office (PRH) — YTJ open data MCP.
CVR — Det Centrale Virksomhedsregister (Danish Central Business Register) MCP.
Nordic company intelligence: look up companies, AI summaries, scores and signals via MCP.
Add one secure layer between your agents and this server.
