cb9eedd7f2fd09b84da9d05ae030cd95be95333f
6
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 MattssonandClaude Fable 5.1
|
26e29f47bc |
feat(company): ideell förening as a third legal form, behind a flag (#2072 step 1) (#2423)
* feat(company): ideell förening as a third legal form, behind a flag (#2072 step 1) Why the problem occurred: the legal form was modelled as a binary flag in ~300 files. `EntityType` was a two-member union, but nothing dispatched on it exhaustively: 28 sites defaulted `?? 'enskild_firma'` (invoice, categorize, match, stripe, invoice-inbox) or `?? 'aktiebolag'` (year-end, bokslut, MCP), and every form-dependent choice was an `=== 'aktiebolag' ? A : B` ternary. Widening the union compiled everywhere and changed nothing, so a förening would have booked as an enskild firma in the app and as an aktiebolag in bokslut and MCP, with no error anywhere. The lookup refused föreningar at the door (mapEntityType returned null), which is what the tester hit. What was removed or simplified: the silent defaults. One module, lib/company/entity-type.ts, now holds the list (ENTITY_TYPES), the parser (never defaults), the resolver (settings hint, then companies.entity_type, then throw) and `byEntityType`, whose Record arms make the compiler refuse the next widening until each site has an answer. The form-dependent facts (closing account, owner settlement account, calendar-year lock, default method, K1/K2 label, personnummer vs 16-prefix) live there once instead of in the ternaries. On the SQL side supported_entity_types() replaces four copies of the literal list in the create RPCs. Why this shape and not the proposed one: the tracker asked for the enum widening plus a chart; that alone was the dangerous version (compiles, books wrong). Bundling stiftelse was considered and dropped: identical plumbing but no chart block. Creation sits behind NEXT_PUBLIC_IDEELL_FORENING_ENABLED so the CHECK, RPCs and seed can ship now and the first partner is switched on without a migration; the flag goes when Phase 2 (packs, INK3, årsbokslut, Swish) lands on the tracker. Domain choices (DECISIONS.md 2026-09-08, verify with an accountant before Phase 2): result closes to 2069 with 2068 as prior-year carry; no owner accounts, member settlement on 2890; accrual default; brutet räkenskapsår allowed; K1 label for the 5 000 kr accrual threshold (BFNAR 2010:1); org number gets the 16 prefix. Migration 20260908110835 widens the three CHECK constraints, adds supported_entity_types(), re-creates the three create RPCs with the widened guard and adds the förening block to seed_chart_of_accounts. Applied to staging and covered by ideell-forening-entity-type.pg.test.ts. Part of #2072 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PdGafpUA7jVV1oYjkwfQCh * fix(company): close the förening paths the skeptic refuted (#2072) Five refutations from the /skeptic pass on 7a05c54d2, each fixed at the shared definition rather than the reported site: 1. Privately paid supplier invoices and the utlägg dialog resolved the owner account in lib/expenses/payer.ts with its own AB/EF ternary, so a förening member's invoice was built on 2893 and then refused by the expense-claim service (which already said 2890), burning an ankomstnummer. The helper now uses ownerSettlementAccount. 2. Booking templates substitute their `_ab` accounts only for an aktiebolag; the `private_expense` template kept its base 2013 for a förening. Template accounts now resolve through templateAccountForForm: EF base, AB override, förening base with owner accounts translated to 2890 (booking-templates.ts and proposal-lines.ts share it). 3. A VAT-registered förening with helårsmoms got no momsdeklaration deadline: the annual VAT rule bailed on anything but AB/EF. A förening is a juridisk person and follows the räkenskapsår schedule (SFL 26 kap 33 §), so the rule now keys on fiscalYearLockedToCalendar instead of the two literals; same in the MCP VAT report. 4. 2069 would have accumulated across years: the year-open omföring was AB-only with 2099/2098 hard-coded. planResultAppropriation now takes the pair from resultClosingAccounts (AB 2099 -> 2098, förening 2069 -> 2068) and skips forms with no carry (EF). 5. With the flag off, a registry lookup that returned "Ideell förening" was prefilled into the onboarding journey, the form picker was skipped and the create step answered "Ogiltig företagsform" with no way back. The journey, the BankID picker, the onboarding page and the MCP lookup now use mapSetupEntityType, which maps only creatable forms, so a flagged-off form falls through to the picker as before. Also: form picker keeps its AB-first order; tests for each fix. Part of #2072 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PdGafpUA7jVV1oYjkwfQCh * chore(migrations): move ideell förening migration after main's latest version (20260908143051) Two migrations landed on main after the branch forked; a lower version would be skipped by the merge-time apply. Staging history row renamed to match. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PdGafpUA7jVV1oYjkwfQCh * chore(skills): regenerate accounted-api reference for the widened entity_type enum Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01PdGafpUA7jVV1oYjkwfQCh --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com> |
||
|
|
ad8566f1ae |
feat(settings): per-company data-analysis opt-in gating the calibration corpus (#1346) (#2007)
* feat(settings): per-company opt-in for data analysis of bookkeeping outcomes (#1346) Adds company_settings.data_analysis_opt_in (default false, no grandfathering) and gates every path that reads bookkeeping outcomes across companies on it: POST /api/agent/categorize/outcome stops writing calibration samples for companies that have not opted in, and the backtest / calibration-fit scripts filter to opted-in company ids. One helper (lib/company/data-analysis.ts) is the single gate for future analysis paths. A toggle on Inställningar > Företag states plainly what is analysed (proposed vs booked account, amount, confidence; no free text, no personal data) in sv and en. The flag is UI-only by design: consent is a human action, so it is absent from the v1 REST / MCP settings pick lists. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015nAd8XJ2RPCmG2eKoLBdna * fix(settings): make data-analysis consent copy true for the backtest path (#1346) Addresses adversarial review findings on PR #2007: - Findings 1-3 (consent narrower than the gated processing): the flag also gates scripts/backtest-categorize.ts, which re-runs transaction descriptions, merchant names and matched underlag through the model. The sv/en toggle help and disclosure now state that explicitly as "evaluation runs" and no longer claim that free text or underlag are excluded. The migration header and COMMENT, the lib/company/data-analysis.ts docstring, the backtest script header and the DECISIONS line say the same. Kept the gate (un-gating would put the script back to reading every company with no consent at all). A test pins that both locales name those inputs and contain no "no free text / no underlag" denial. - Finding 4 (member sees an active switch that RLS rejects): the toggle is now enabled only for owner/admin, matching the company_settings update policy; the disclosure says only administrators can change the choice. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015nAd8XJ2RPCmG2eKoLBdna * fix(scripts): address round-2 review findings (#1346) 1. [minor] Opted-in company filter was an unbounded PostgREST `in` list in the URL (scripts/fit-categorize-calibration.ts, scripts/backtest-categorize.ts). Both scripts now read the opted-in ids through a shared, paginated helper (listDataAnalysisOptedInCompanyIds, fetchAllRows so the pre-fetch no longer caps at 1000) and query per chunk of 100 ids (chunkCompanyIds). The fit script pages each chunk on the id PK; the backtest merges per-chunk results and re-cuts to the N most recent overall. Early exit on zero opt-ins is kept. Pinned with tests in lib/company/__tests__/data-analysis.test.ts. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015nAd8XJ2RPCmG2eKoLBdna * fix(scripts): coerce a null transaction description in the backtest (#1346) The typed row from the chunked consent query made description nullable, which TransactionForSelect does not accept; fall back to the original description or an empty string, as the untyped row did implicitly before. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015nAd8XJ2RPCmG2eKoLBdna --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
1a27b5bd4a |
fix(auth): route document integrity, transaction delete/list and agent categorize through withRouteContext (#1926)
Two handlers hand-rolled supabase.auth.getUser() and therefore skipped the MFA (AAL2) gate on hosted: DELETE /api/transactions/[id] and GET /api/transactions. Both sat next to a sibling handler that was already wrapped, and the raw-route-auth ratchet exempted a file as soon as any withRouteContext call appeared in it, so they were never flagged. GET /api/documents/[id]/integrity and POST /api/agent/categorize called requireAuth() directly (MFA enforced, but no request id, no completion log, no canonical error envelope). All four are now withRouteContext handlers with identical company scoping and responses; the transaction delete keeps its viewer rejection via requireWrite. The guard now judges each top-level export segment of a route file on its own, so a wrapped handler no longer exempts a hand-rolled sibling. Baseline is unchanged (mcp-oauth/authorize remains the one grandfathered file). Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
72c81c21e7 |
feat(categorize): feed the selector the underlag, not just the bank line (#1785)
The highest-leverage quality lever for real users. A prod read showed the majority are cold-start (365 companies, 32.7k unbooked transactions, median 0 counterparty templates), so the LLM selector carries them — and it was only seeing the bank line (merchant + amount), never the receipt. - lib/agent/categorize/underlag.ts: gathers the matched receipt/invoice text for a transaction (receipts.matched_transaction_id + invoice_inbox_items .matched_transaction_id + the transaction's own attached document) and renders it as bounded Swedish text — supplier, date, total, moms, line items. Same sources the categorization intent reads, as a string not a tool loop. Core queries the tables directly (no @/extensions import). Best-effort: '' on any failure. - POST /api/agent/categorize gathers it server-side when the caller didn't supply `underlag`, so the model reasons over the actual supplier + line items. Server-side only, no client change. 31 categorize tests green; lint + guards + scoped typecheck clean. Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
|
704bf93e08 |
feat(categorize): confidence calibration engine + measurement loop (cascade step 4) (#1784)
Turns the selector's raw confidence into a score that means what it says. - lib/agent/categorize/calibration.ts: the engine. Isotonic regression (pool-adjacent-violators, distribution-free + monotonic) over (confidence, was_correct) samples → a calibrator; plus reliabilityByBucket, ECE, and bandFor(). bandFor NEVER returns 'auto' without a fitted calibrator (no silent booking on an unproven score) and never auto-books above an amount cap. 12 engine tests (overconfidence pulled down, underconfidence lifted, monotonicity, ECE, band gating). - Measurement loop: migration categorize_calibration_samples (append-only, company-scoped RLS, confidence CHECK [0,1]) + POST /api/agent/categorize/ outcome logging one sample (proposed vs actually booked) fire-and-forget from QuickReviewDialog on a successful book (sandbox skipped). AiCategorizeProposal surfaces the proposal metadata via onProposal. - scripts/fit-categorize-calibration.ts (read-only): prints the reliability diagram + ECE + fitted calibrator once data has accumulated. Fitting needs a few hundred real outcomes, so nothing calibrates today — the loop starts collecting, and "säker" stays uncalibrated (no auto-book) until the data proves it. 131 unit tests green; RLS covered by a pg-real test. Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
|
e7a5e65ecf |
feat(categorize): Tier 1 candidate gathering + the proposal route (#1781)
The auto-booking cascade end to end (retrieval → selector), minus the write. - lib/agent/categorize/candidates.ts (Tier 1): assembles the deterministic candidate slate for a transaction — the learned counterparty template (strongest, carries its own VAT) plus mapping rules / patterns / per-merchant history via the same engine gnubok_suggest_categories uses. No model call. Deduped by account (highest confidence wins), capped; suggestions get the category's default VAT treatment derived. - POST /api/agent/categorize: loads the transaction + company VAT context, runs Tier 1 → Tier 2 selectAccount, returns the proposed account + VAT + confidence + reasoning + the candidate slate. Never posts anything — the caller renders an approval card. Gated on configured (any provider incl. local), same gates as /api/agent/ask. 12 tests: candidate merge/dedupe/VAT-derivation, and the route (401/429/400/ 403/404/503 + happy path threading entity type, VAT, underlag, samples). Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |