Commit Graph

606 Commits

Author SHA1 Message Date
Jonas Flodén 8094f288fa fix(ci): quote workflow name with embedded colon in compliance-swarm.yml (#898)
The unquoted "compliance: full swarm" value contains a colon+space,
which YAML parses as the start of a nested mapping, causing GitHub
Actions to reject the whole file as invalid.

Signed-off-by: Jonas Flodén <jonas@floden.nu>
2026-07-06 12:08:42 +02:00
Jakob Wennberg 9b126d22b9 fix(reconciliation): server-side confidence floor for unattended auto-apply (#903)
* fix(reconciliation): server-side confidence floor for unattended auto-apply

runReconciliation applied every greedy match, including auto_fuzzy at
confidence 0.75, with no server-side threshold. The UI is checkbox-gated,
but the unattended callers (enable-banking nightly sync cron, the
extension's post-sync sweep, and the v1 run endpoint) had no guardrail.

- Add ReconciliationOptions.confidenceThreshold (0..1, clamped): the
  apply loop skips matches below it. Skipped matches stay in the result's
  matches array and are counted in the new skippedBelowThreshold field,
  so they are reported for review rather than silently dropped. Dry runs
  are unaffected; omitting the threshold preserves current behavior.
- Both enable-banking sync callers now pass
  DEFAULT_UNATTENDED_CONFIDENCE_THRESHOLD (0.9, mirroring the
  gnubok_auto_match_period MCP default), so unattended runs never commit
  fuzzy (0.75) or date-range (0.85) matches.
- v1 POST /reconciliation/bank/run accepts confidence_threshold
  (optional, 0..1, mirroring the MCP tool naming) and returns
  skipped_below_threshold; registry docs/pitfalls updated.

Closes #880

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(reconciliation): surface skippedBelowThreshold in unattended sync logs

Review finding on the first pass: the floor's 'reported, not silently
dropped' guarantee never reached the two unattended callers, which
discarded or under-logged the result. Also logs the DECISIONS.md line
for the deliberate no-default choice on the v1 route.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-06 11:30:46 +02:00
Jonas Flodén b700108107 feat(mcp): add invoice_date_override to create_supplier_invoice_from_inbox (#848)
Unblocks inbox items where OCR missed the invoice date: priority is override, then extracted invoiceDate, then the existing hard error. Both date overrides now reject non-ISO input at staging time instead of failing opaquely at booking. Includes dry_run and validation tests.
2026-07-06 11:24:05 +02:00
Jakob Wennberg ac1529e413 fix(ui): block illegal VAT rates, honest opening-balance state, surface account-save errors (#902)
* fix(ui): block illegal VAT rates, honest opening-balance state, surface account-save errors

- Supplier invoice form (#863 item 1): onSubmit now blocks any
  non-reverse-charge line whose VAT rate is outside the legal Swedish
  set {25, 12, 6, 0} with a destructive toast naming the line and the
  legal rates. Server-side schema tightening stays out of scope.
- Opening balance import (#837): the summary derives isBalanced from
  the running totals (0.01 epsilon) and renders the AlertCircle
  destructive pattern with the differens amount when unbalanced;
  canExecute includes isBalanced so the commit button is disabled
  instead of funneling users into a server-side rejection.
- EditAccountDialog (#838): a failed account save now shows a
  destructive toast with the server-provided message instead of being
  silently swallowed; the dialog stays open for retry.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(ui): compare opening-balance totals in whole ore, map account-save errors to Swedish

Review findings on the first pass: the < 0.01 epsilon misclassified exact
1-ore imbalances as balanced (0.03 - 0.02 evaluates just under 0.01), and
the save-failure toast surfaced raw English server text instead of routing
through getErrorMessage like the sibling handlers in the same file.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-06 11:20:46 +02:00
Jakob Wennberg 8e7e7201d3 fix(db): drop delete_user_account RPC that bypassed BFL retention (#901)
* fix(db): drop delete_user_account RPC that bypassed BFL retention

delete_user_account disabled the retention/immutability/audit triggers,
deleted audit_log rows, and cascaded auth.users, destroying 7 years of
legally retained rakenskapsinformation (BFL 7 kap 2 paragraf). It was
SECURITY DEFINER with only a self-only guard and no REVOKE, so any
authenticated user could call it via PostgREST.

The product path already uses anonymize_user_account, which so far
existed only on production (drift). This migration drops the dangerous
RPC, commits the prod definition of anonymize_user_account verbatim,
adds the profiles tombstone columns it writes (also drift), and locks
grants down to authenticated only.

Closes #342

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs: log profiles tombstone-column drift-capture decision

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-06 11:20:34 +02:00
Jonas Flodén b5f568004f fix(ci): fail loud when the compliance diff artifact is missing (#832)
Follow-up to #830: the review script now throws if DIFF_FILE is set but the artifact file is missing, instead of silently reviewing a base-vs-base diff and posting a misleading 'No diff detected' comment. The fallback filename parser also captures deleted files, and the Bedrock job gets a 10-minute timeout.
2026-07-06 09:29:01 +02:00
Jonas Flodén f58b7a3602 fix(mcp): treat created_journal_entry_id as a terminal link in gnubok_list_unmatched_documents (#891)
Documents booked directly as a journal entry (e.g. via SIE import) no longer appear as unmatched. Aligns the tool with the terminal-link predicate already used by gnubok_list_inbox_items and lib/worklist, preventing agents from double-processing already-booked documents. Adds a regression test suite.
2026-07-06 09:27:51 +02:00
Jakob Wennberg c43c4a076c feat(salary): allow recalling approval on a salary run (approved → review) (#894)
* feat(salary): allow recalling approval on a salary run (approved → review)

An approved run was a dead end: the only forward path was paid → booked,
so a wrong salary snapshot (e.g. stale employee monthly pay) could not be
fixed without paying and then storno-correcting. Approval is an internal
control point — nothing legally binding happens until payment, booking,
or AGI filing — so recalling it is allowed until the AGI reaches
Skatteverket.

- POST /api/salary/runs/[id]/unapprove: approved → review; clears
  approved_by/at and payment-file tracking; deletes generated-but-unfiled
  AGI declarations (stale XML must not stay exportable); 409 once the
  AGI is pending_signature/submitted/accepted — correction AGI (same
  specifikationsnummer) is the lawful path then.
- New salary_run.approval_reverted event for the audit trail.
- "Ångra godkännande" secondary action on the run page with a
  consequence-aware confirm (payment file possibly at the bank, sent
  payslips, generated AGI), sv + en.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(salary): delete stale AGI after the unapprove transition, not before

Bot-review triage on #894: the declaration delete ran before the
optimistic status update, so a failed transition (concurrent flip,
transient error) would have destroyed the generated AGI while the run
stayed approved. Flip the run first; a delete failure afterwards is
harmless (agi_generated_at is already null, regeneration upserts over
the orphan). Also record the deleted declaration id in the
approval_reverted event payload, and warn in the confirm dialog that a
manually filed AGI requires a correction declaration instead.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(salary): close the unapprove TOCTOU on concurrent AGI filing

Superagent P2 + compliance-bot round 2 on #894: AGI submission is
allowed from approved (also out-of-band via MCP/public API), so a
filing could land between the route's read and its update, and the
route would flip the run and delete a submitted declaration.

- Re-assert agi_submitted_at IS NULL inside the optimistic update
  filter, not just on the stale read.
- Guard the declaration delete with the same status filter so it
  no-ops if the declaration advanced since the read; log a miss.
- Zero-row update (PGRST116) now returns 409 "status har ändrats"
  instead of a generic 500.
- The approval_reverted event only reports deletedAgiDeclarationId
  when a row was actually deleted.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 19:44:22 +02:00
Mattsson 2c2743eb79 Check/salary bankid api (#892)
* fix(bankid): harden login/signup flow — polling, signup rollback, metadata merge, enrichment lookup

- middleware: read BankID enrichment from the bankid_enrichment table (the
  extension_data path has been dead since the multi-tenant refactor), so
  company-less BankID users land on /select-company instead of the manual wizard
- BankIdAuth: hard 6-min poll deadline; every failed poll counts toward the
  give-up limit; guard overlapping ticks so completion runs exactly once
  (a double /complete regenerated the magic link and invalidated the first,
  failing logins intermittently); retry clicks wait out the start cooldown
  instead of silently no-oping; Swedish messages for 429/unknown start errors
- bankid/complete: all-or-nothing signup — delete the created user when the
  identity insert, app_metadata update, or magic-link generation fails, so a
  retry starts clean instead of hitting account_exists with an unusable account
- bankid/unlink: read-merge-write app_metadata so has_password survives unlink
  (BankID-only users could otherwise strand themselves with no login method)
- login: BankID "create account" CTA now links to /register instead of
  dismissing the notice; sv.json: fix missing å/ä/ö in settings_bankid strings

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs: move secondary guides into docs/, delete dead root files

Move DOCKER.md, SELF-HOSTING.md, WHITELABEL.md and extensions.md
(renamed EXTENSIONS.md) into a new docs/ folder and update all path
references (README, setup.sh, .dockerignore image rules, docker-publish
workflow comment, _example-branding, lib/branding/service.ts).

Delete two dead root files: customer.json (stray API-test payload) and
findings.md (point-in-time swarm audit export, criticals already filed).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Signed-off-by: Emil <emilmattsson14@gmail.com>

* fix(api): security & correctness hardening + withRouteContext MFA migration across API routes

Audit of ~100 app/api routes. Highlights:

Security
- agent/conversations: list leaked colleagues' titles + message previews
  (company-scoped RLS, no user filter) -> user-scoped
- calendar/feed PUT: raw body into .update() allowed feed_token fixation on a
  public unauthenticated URL -> strict schema, content toggles only
- bokslutsdispositioner: unbounded schablonintaktRate could inflate the
  IL 30 kap 25% periodiseringsfond cap base -> bounded
- agent profile/composer/onboarding: viewers could rewrite the agent profile
  while sibling /verify blocked them -> role-gated

Correctness
- account-totals / listAssets: unbounded queries silently truncated at 1000
  rows (under-counted money; skipped assets at year-end depreciation) ->
  fetchAllRows with stable order (+3 more pagination fixes)
- voucher-gaps: swallowed detect_voucher_gaps RPC errors (BFNAR gap view could
  show "no gaps" when the check never ran) -> surfaced
- 5 phantom-success writes (OK on zero matched rows) fixed
- assets K3 component-sum validated against stale acquisition_cost -> fixed
- invite silent email-send failure -> response carries email_sent;
  deadlines/calendar cast-then-check JSON crashes -> Zod

Convention
- ~44 legacy routes converted to withRouteContext (MFA); added Zod validation,
  corrected status codes, console.* -> lib/logger

Response shapes preserved for existing callers. ~110 new tests.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(bookkeeping): save a booking as a reusable template from Bokför direkt

Add a "Spara som mall" action to the manual booking dialog so users can
capture a kontering they just worked out as a booking template — right
where they figured out how something should be booked.

- derive amount-parameterised template lines from the concrete booking
  (settlement = the non-VAT leg nearest the total, 26xx = a VAT line with
  its rate snapped to the nearest standard rate, the rest = business
  ratios; line labels come from the loaded BAS chart)
- extract the shared TemplateForm out of BookingTemplatesPanel so the
  booking dialog reuses the same editor, live preview and convertibility
  hints instead of duplicating them
- save via the existing POST /api/settings/booking-templates endpoint

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(bokslut): render arsredovisning RR/BR at ÅRL post level — no kontonummer

Bolagsverket rejected a user's filed årsredovisning with "Balansräkning
och resultaträkning ska inte innehålla kontonummer": the PDF built every
statement row as per-account "1930 Företagskonto" lines while the iXBRL
filing path already aggregated to statutory posts, so the two artifacts
diverged.

The PDF statements now derive from the same K2 risbs mapping the iXBRL
document uses (mapTrialBalancesToK2), via a new statement-rows.ts that
emits post-level rows in uppställningsform order for both the K2 and K3
templates. Also fixed along the way:

- Jämförelseår column (ÅRL 3:5 §) — previous-year trial balances now load
  and render; the old PDF had no comparatives at all.
- mapping.warnings (unmapped accounts, RR ≠ 2099, obalans, reclass
  nudges) flow into ArsredovisningData.warnings so the wizard flags a
  non-fileable document before download.
- Flerårsöversikt current/previous year overridden with the mapper's
  strict-3000–3799 Nettoomsattning, mirroring build-input's
  duplicate-fact rule, so the FB table ties to the RR.
- FB eget kapital-table is post-level and drops obeskattade reserver
  (never eget kapital); K3 equity-changes statement uses real prior-year
  opening balances with derived utdelning/nyemission residuals that tie
  the roll-forward exactly to booked UB.
- build-input dedupes warnings now that the PDF path runs the same
  mapping.

Regression test asserts no RR/BR label ever contains a four-digit
account number again.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(reports): diagnose untransferred prior-year results behind balance-sheet differens

Prod incident (97 kr): a multi-year SIE migration lacked one year's
omforing av arets resultat; the residual corrupted every later derived
opening balance and Balansrakningen showed a bare "Differens: 97 kr"
with no explanation. Continuity checking cannot catch this failure mode
(prior-year UB and derived IB match per-account by construction) - the
invariant that actually breaks is per-year P&L = 0 for all non-latest
years.

- lib/reports/imbalance-diagnosis.ts: shared detector
  (findUntransferredResults + buildImbalanceDiagnosis)
- Balansrakning/Balansrapport attach imbalance_diagnosis when unbalanced,
  naming the exact culprit years; rendered in web views + PDF; MCP
  gnubok_get_balance_sheet inherits the field via spread
- SIE import: parse-time warning when a completed year's vouchers leave
  a P&L residual, plus a post-import DB walk surfacing culprits as
  warnings and structured details.untransferredResults; the Arcim
  migration workspace previously dropped result.warnings entirely and
  now renders them
- opening-balance/correct: pre-flight the company lock date and return
  409 OB_COMPANY_LOCK_DATE (retryable: false, lock date interpolated in
  the client message) instead of the retryable 500 that invited blind
  retries; catch-path maps a raced trigger rejection to the same code

Diagnosis runs only on unbalanced paths (zero cost when healthy) and
never fails the report or the import. No migration, nothing persisted.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix: production error remediation — FX rates, deadlines, log levels, correction relink

Batch of fixes for recurring Vercel runtime errors:

- Riksbanken FX rates: persistent read-through cache (exchange_rates
  table), one retry honoring Retry-After on 429/5xx, bounded ingest
  concurrency, and an honest fallback — most recent cached observation
  or null, never a hardcoded rate silently booked into amount_sek.
  Unrated transactions stay repairable via refresh-exchange-rate.
- Tax deadline regeneration inserts replacement rows before deleting
  the superseded set, so a failed insert no longer wipes a company's
  deadlines (the 23502 user_id regression did exactly that). Migration
  makes deadlines.user_id nullable for system-generated rows.
- Route wrappers + errorResponse log 4xx outcomes at warn so only
  genuine 5xx reach Vercel's runtime-error clustering; client-supplied
  /api/log telemetry demoted to warn as well.
- application/json documents (raw PSD2 responses archived per BFL)
  validate as parseable JSON with object/array root instead of always
  failing the magic-byte check.
- correctEntry surfaces document-relink failures to callers, and the
  BFL document-immutability trigger now allows relinking underlag from
  a reversed entry to its correction (migration + pg test).
- Middleware clears stale session cookies on /api requests too, using
  scope 'local' so cleanup doesn't re-trigger the failed token refresh.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(skatteverket): persist token health and stop retrying dead consents

Terminal auth errors (SESSION_EXPIRED, REFRESH_EXHAUSTED, MISSING_SCOPE,
TOKEN_CORRUPTED) mark the token row needs_reconsent with the error code
and timestamp — SKV per-flow refresh tokens live 65 minutes, so once
expired nothing recovers without a fresh BankID consent. The AGI
kvittens and skattekonto sync crons skip flagged connections instead of
failing every night, and the settings panel prompts for re-consent
proactively. A successful reconnect resets the row to active.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(banking): allocate distinct BAS ledger slots for PSD2 mirror accounts

A bank returning N same-currency accounts used to map them all onto the
currency default (1930/1932/1933/1934), tripping the UNIQUE
(company_id, ledger_account) constraint per-account — swallowed errors
left accounts silently unmirrored. allocatePsd2LedgerAccount now hands
out the currency default first, then free 1931–1959 sub-account slots,
skipping slots held by any existing row.

- Callback persists allocations to accounts_data so the picker pre-fills
  reality; reconnect reuses previously mirrored ledgers instead of
  re-deriving (a user remap to 1935 survives).
- Selection save resolves effective ledgers up front and rejects
  duplicates or cross-connection conflicts with a 400 instead of
  silently skipping the mirror.
- Bank error codes + psu_type are forwarded to the settings page for
  every OAuth error, keying the Handelsbanken corporate fullmakt
  guidance.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(agent): stage exact journal lines on categorization previews

Categorization previews only carried debit/credit accounts, the GROSS
amount, and separate VAT rows — read together that looks like an
unbalanced 'gross on cost account + VAT debit' entry, and it misled
both users and agents into rejecting correct proposals. The MCP
preview and the pending-operation PATCH now materialize the exact
lines the commit executor will post (net cost line, VAT line, gross
bank line, SEK) via buildTransactionEntryLines, and PATCH re-derives
them from the new mapping instead of spreading stale staged lines.
ApprovalCard and /pending render the verifikat lines, falling back to
the legacy summary only for operations staged before this fix.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* feat(bookkeeping): prune unused imported accounts from the chart

SIE imports routinely bring in hundreds of accounts that were never
used and clutter the kontoplan. New account_usage_counts RPC (one
grouped query instead of a count per account) backs GET
/api/bookkeeping/accounts/usage, and POST /api/bookkeeping/accounts/prune
deletes zero-usage accounts — dry-run first, then an explicit account
list capped at 2000. Accounts with journal lines are skipped, never
deleted. The chart manager shows a usage column and a prune dialog
grouping custom accounts vs unused BAS-seeded ones.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* feat(api): carry dimensions through v1 invoice and supplier-invoice surfaces

Credit-note creation now copies default_dimensions and per-line
dimensions from the original, so the reversing journal entry nets
against the same dimension cells instead of dropping them. List/detail
responses expose the dimension fields, and the OpenAPI spec snapshot
follows.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* perf: batch serial Supabase round-trips on hot dashboard paths

Every dashboard render pays the layout's query chain, so serialized
awaits are direct wall-clock: the layout, chat conversation, invoice
detail, supplier detail, select-company, and agent-onboarding pages now
run their independent lookups in parallel batches, and
getCompanyCapabilities folds its disabled-config read into the same
round-trip. JournalEntryList hydrates the saved fiscal-year scope
optimistically instead of serializing the first entries fetch behind
the fiscal-periods request. The supplier detail page filters invoices
server-side via a new supplier_id query param instead of fetching the
whole company ledger, and the invoice editor (with its framer-motion
dependency) lazy-loads so it stops shipping with the invoice list
bundle.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* feat(salary): one-click runs, payslip delivery, payments settings, run cockpit

Salary P1 batch, driving the 20-click flow toward 3 clicks:

- One-click 'Starta lönekörning': POST /api/salary/runs accepts an
  empty body and resolves defaults server-side — period follows the
  latest non-corrected run, payment date from the new
  salary_pay_day setting, series from the per-source-type map. The
  separate /salary/runs/new page is gone.
- Run detail page rebuilt as a step-railed cockpit (progress rail,
  KPI cards, employee ledger, journal preview) on a deliberately
  wider canvas; components extracted to components/salary/run/.
- Payslip delivery: tokenized public payslip pages (/payslip/[token],
  backed by salary_payslip_links) plus per-employee email send with
  PDF — employees need no account, and the middleware exempts the
  route from auth redirects.
- Payments settings: salary pay day, default bank, and pain.001 vs
  Bankgirot Lön format with per-bank upload instructions and an LB
  sunset warning (banks retire LB during 2026).
- AGI panel: full submission status flows (stale drafts, signing
  links, kvittens polling, error reports); tax payment panel with
  skattekonto shortcut and mark-as-paid.
- Salary calendar bulk editing, employee benefits/tax-card polish,
  municipality tax-table lookup improvements.

messages/sv+en also carry the strings for the account-prune,
skatteverket-reconsent, and banking surfaces committed just before
this.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* chore: adopt Next 16 proxy.ts convention + repo housekeeping

- Rename middleware.ts to proxy.ts with the proxy() export (Next 16
  renamed the middleware convention; behavior unchanged).
- Exclude dev_docs/ from tsconfig so stray snippets in planning docs
  don't break the build type-check.
- Ratchet antipatterns-baseline down (raw-route-auth 165 → 119) to
  lock in the withRouteContext migration from 5cfd2b76.
- template-library uses roundOre() instead of inline rounding.
- database.md: drop account_balances from the key-tables list.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(bookkeeping): robust service-role detection in correction document relink

relink_documents_to_correction() keyed its service-role branch on auth.role(),
which reads the singular request.jwt.claim.role GUC that PostgREST v10+ and the
pg-real harness no longer populate. Genuine service-role callers (pending-ops
executor / MCP approve) landed in the auth gate and could not relink underlag.
Read the role from the request.jwt.claims JSON directly, mirroring the canonical
link_voucher_rpcs_tenant_guard convention. Validated on staging.

Also: harden the salary run page's error paths (res.json().catch) against
non-JSON error bodies, and roll back the pg-real service-role case in finally so
an aborted transaction cannot poison a pooled connection for the next test.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(documents): restore journal_entry_line_id link durability (BFL 7 kap)

Migration 20260704103000 rewrote enforce_document_journal_entry_immutability to
guard journal_entry_id but left journal_entry_line_id to the metadata trigger,
which exempts draft-linked docs -- and the entry-level trigger only fired on
UPDATE OF journal_entry_id, so a line-id-only UPDATE never invoked it at all.
That let a set journal_entry_line_id be cleared to NULL, breaking the "link
durable from first set" invariant (document-immutability.pg regression).

Widen the trigger to fire on journal_entry_line_id too and guard it with the
same uuid-durability rule as journal_entry_id (setting NULL -> uuid stays
allowed; clearing/re-pointing a set value is blocked, status-independent). The
correction-relink GUC path, which legitimately clears line_id when moving
underlag to the posted correction, stays exempt. Validated on staging.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Signed-off-by: Emil <emilmattsson14@gmail.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-05 03:05:09 +02:00
Jakob Wennberg ec27228a8e style: remove em/en dashes repo-wide, add CLAUDE.md rule against them (#890)
Em dashes (—) and en dashes (–) had spread across comments, docs, tests,
and a few UI strings, reading as AI-generated boilerplate rather than
house style. Replaced each with punctuation matching its context: colon
for explanatory clauses, comma for asides, plain hyphen for numeric/legal
ranges (e.g. "21-23§"), "to"/"till" for date ranges, parentheses for
paired-dash asides. messages/en.json and messages/sv.json were fixed by
hand together to keep sv/en in sync.

Left untouched where the dash is the functional subject rather than
decorative punctuation: date-range-parser.ts's separator regex,
charset-repair.ts's CP1252 byte-mapping table (and its test), the SIE
encoding mojibake docs, generic-csv.ts's minus-sign normalizer, the
agent system-prompt files that already instruct against em dashes, and
a golden iXBRL test fixture compared byte-for-byte.

Also fixes two bugs surfaced along the way: an off-by-one in
ApiKeysPanel's scope-label split (a leftover from an earlier partial
pass), and a charset-repair test that had lost the literal en-dash it
exists to verify.

Regenerated the agent atom seed migration (skills:generate) since 27
SKILL.md files changed. Added a CLAUDE.md rule against em/en dashes,
with an explicit carve-out for the functional-dash cases above.

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-04 15:58:06 +02:00
Jakob Wennberg a5154ee884 docs(dimensions): post-merge review nits from #888 (#889)
- SIE round-trip fixture: vouchers in ascending verno order (A1, A2, A3)
  per SIE4 core invariant 5 — the custom-dim voucher was spliced out of
  order
- commitEntry: note that source_type is a HEADER column repeated by the
  join — reading lines[0] IS reading the entry header, lines cannot mix
  source types (a reviewer misread this as per-line logic)
- dimension-rules: sharpen the credit-note exemption rationale — credit
  notes copy the original's bags, so enforcement is either a no-op or
  would force the exact asymmetric-tag P&L skew the feature prevents

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 17:26:00 +02:00
Jakob Wennberg 86c924a6e9 fix(dimensions): exempt system source types from account dimension rules (#888)
SIE import books three system entries through the engine (opening
balances, IB resynk, the omföring adjustment for excluded vouchers) —
after PR10 those passed through the rules layer, so a required rule
could block an import and default/fixed rules could inject dimensions
into derived historical entries. Year-end, currency revaluation and
credit instruments had the same exposure.

Policy now governs NEW business events only: source types in
DIMENSION_RULE_EXEMPT_SOURCE_TYPES (opening_balance, import, year_end,
storno, correction, credit_note, supplier_credit_note,
currency_revaluation, system) skip both the draft-time apply and the
commit-time assert — imported history lands verbatim (BFL 5 kap),
bokslut can never be blocked by a dimension rule, and crediting an
entry that pre-dates a rule always works. Operational sources (manual,
bank_transaction, invoice_*, supplier_* registrations/payments,
salary_payment) stay enforced.

The SIE round-trip test now also covers a PR10-created custom dimension
(#DIM 20) with a custom child (#UNDERDIM 25 ... 20) and a tagged line —
proving user-created dims survive export → parse structurally.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 17:17:16 +02:00
Jakob Wennberg 764348e99c feat(dimensions): PR10 advanced — custom dimensions, hierarchy, account rules, commit enforcement (#886)
* feat(dimensions): PR10 advanced — custom dimensions, hierarchy, account rules, commit enforcement

The final rung of the dimensions ladder
(dev_docs/dimensions_implementation_plan.md §7 row 10):

- custom dimensions: POST /api/dimensions creates registry dims (next free
  SIE number >= 20 when omitted; explicit numbers allowed — SIE import
  already mints reserved ones); register gets a 'Ny dimension' dialog with
  a quiet Avancerat disclosure for the #UNDERDIM parent; GET now carries
  parent_sie_dim_no (the column + SIE round-trip existed since PR1/PR5 —
  this exposes it)
- account_dimension_rules (migration 20260703120000): one rule per
  (account, dimension) — required / default / fixed, per-rule is_active,
  company-scoped RLS, composite FK to the registry, value-presence CHECK
- enforcement, opt-in BY CONSTRUCTION (zero rules = engine byte-identical;
  deliberately NO settings toggle — a rule that exists but is ignored is
  worse than either extreme): default/fixed apply onto line bags at draft
  creation (fixed overwrites, default fills); required asserts at
  commitEntry with a Swedish MANDATORY_DIMENSION_MISSING naming every
  account + dimension; the bulk-book route runs the same policy before its
  RPC; storno/correction paths never pass through commitEntry so history
  always reverses regardless of policy; rule fetches fail open incl.
  thrown exceptions
- chart of accounts: per-account Dimensionsregler section in
  EditAccountDialog (Krävs/Förval/Låst, value picker, pause switch),
  gated on the existing dimensions toggle, quiet when empty
- pickers: LineDimensionFields is registry-driven (one combobox per active
  dimension, cached fetch, hardcoded 1/6 fallback) — every existing mount
  lights up custom dims with zero changes
- agent briefing: per-dimension required_on_accounts/default_on_accounts
  so agents self-correct instead of bouncing off the policy error
- rules CRUD API with existence/active/company validation and qualified
  DTO ids; firm_id FK deferred until the firms table lands (per plan)

39 new tests (pure-fn rules, engine enforcement, both new API surfaces,
pg-real RLS/CHECK/cascade suite); full suite 6,791 green; migration
replayed on a fresh container.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: renumber migration to 20260703200000 — version collision with prod

The concurrent session shipped pending_operations_add_link_document_to_voucher
as 20260703120000 today; the Supabase preview branch (cloned from prod)
rejected the duplicate version key.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: review round — auto-pick retry on collision, fail-open warnings, query schema

- POST /api/dimensions retries once past a concurrent number claim when the
  number was auto-picked (explicit choices still 409)
- every fail-open skip of the dimension-rules policy now logs a structured
  warning (engine draft/commit paths + bulk-book) — deliberate fail-open,
  but observable
- GET /api/dimensions/rules validates its query through
  ListDimensionRulesQuerySchema instead of an inline regex

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 16:50:28 +02:00
Jakob Wennberg 70e893b8d4 fix(skattekonto): stop conflating live-call auth failures with "inte anslutet" + surface SKV connection first in tax settings (#887)
The sync endpoint returns 401 for six distinct auth states (handleSkvError):
NOT_CONNECTED, SESSION_EXPIRED, REFRESH_EXHAUSTED, MISSING_SCOPE,
TOKEN_CORRUPTED, TOKEN_REVOKED. The page treated every 401 as "Skatteverket
är inte anslutet" — directly contradicting Inställningar, which reads the
stored token metadata and truthfully shows "Ansluten" for all but the first.

- syncNow now checks the structured error code: only NOT_CONNECTED flips to
  the not-connected empty state. Every other auth failure renders the
  server's actual Swedish message as a persistent banner with an "Anslut
  igen" CTA, keeping the stored saldo/transactions visible.
- SkatteverketConnectPanel moves to the top of /settings/tax — the
  skattekonto and momsdeklaration pages send users there specifically to
  (re)connect, and below the tax form it sat out of view.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 16:35:09 +02:00
Jakob Wennberg 9f7c842a33 feat(skatt): setup gates + auto-loading momsdeklaration across Skatt & bokslut tabs (#885)
* feat(skatt): setup gates + auto-loading momsdeklaration across Skatt & bokslut tabs

Every tab in the Skatt & bokslut nav group now tells an unconfigured user
what is missing and where to fix it, instead of dead-ending or rendering
zeros:

- Momsdeklaration: gates on vat_registered with a settings CTA; auto-fetches
  the configured period on load and on every period change (no more "Hämta"
  button); one period control (räkenskapsår picked inline for helårsmoms,
  shell selector + back link dropped via new ReportDescriptor.standalone);
  raw <select>s replaced with the Select primitive; banner when moms_period
  is missing; BankID connect returns to the page instead of the report
  library.
- Deadlines: callout (sv+en) when no system-generated tax deadlines exist —
  they are derived from tax settings, so point at /settings/tax rather than
  presenting an empty manual todo list.
- Årsbokslut: "no räkenskapsår yet" (CTA to bookkeeping settings) is now
  distinguished from "nothing to close yet".
- Skattekonto: non-auth fetch failures render an error card with retry
  instead of the misleading "inget saldo hämtat ännu" empty state.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(skatt): address bot review — res.ok guard, blocking moms_period gate, panel hidden while räkenskapsår unresolved

- Auto-fetch treats non-2xx or unparsable responses as errors (with retry)
  instead of rendering undefined data.
- momsPeriodMissing now blocks the declaration (EmptyState + settings CTA)
  rather than fetching a guessed quarterly period behind a banner — a
  declaration submittable for the wrong period type is a hazard, not a
  convenience.
- SkatteverketPanel is not rendered while yearly mode awaits a fiscal
  period, so its actions can never target an unconfirmed period.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 16:12:51 +02:00
Jakob Wennberg 100a4d1291 feat(ux): create salary runs, employees & recurring schedules in modals (#883)
* feat(ux): create salary runs, employees & recurring schedules in modals

The last three full-page create flows move to ?new=1 URL-driven dialogs,
matching the verifikat/invoice pattern (#861):

- Salary run: 4-field form on /salary — creation was pure interruption
  before landing on the run-detail workspace.
- Employee: the last register entity still page-based after customers,
  suppliers, and articles.
- Recurring schedule: consistency with the invoice modal it feeds.

Old /new routes survive as redirects so bookmarks and agent intents keep
working. Dialogs close explicitly (header X / Avbryt) so half-typed forms
survive stray Escape or backdrop clicks.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs: move DECISIONS.md to repo root

dev_docs/ is gitignored, so the decision log was invisible to other
developers. Root matches the existing convention (CONTRIBUTING.md,
SECURITY.md). CLAUDE.md pointer updated.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(lint): ignore Claude Code worktrees in eslint walk

.claude/worktrees/ holds full repo copies; without the ignore, local
npm run lint / check:lint walks them until the ratchet's 64 MB JSON
parse buffer overflows. CI is unaffected (no worktrees there).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 15:18:20 +02:00
Mattsson 237b77a366 feat: custom inbound mail domains, rot/rut payout file, invoice email texts, security hardening (#878)
* fix(security): guard MCP test keys, RLS role gate + voucher RPC guards, /api MFA gate, deps

- MCP: force dry-run / block writes for test-mode API keys in tools/call (extensions/general/mcp-server)
- DB: current_user_can_write role gate on write policies (40 tables) + tenant guards, SET search_path, REVOKE anon on commit_journal_entry / next_voucher_number / detect_voucher_gaps (migration 20260702093000)
- Middleware: MFA (AAL2) gate on cookie-authenticated /api routes via apiPathSkipsMfaGate
- Deps: npm audit fix clears mailparser/linkify-it/nodemailer/svix/uuid highs; xlsx -> SheetJS 0.20.3

Adds unit + pg-real tests. Does not touch in-progress ROT/RUT or invoice-email-texts work.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(invoices): rot/rut begäran om utbetalning — HUS XML (V6), payout tracking + settlement, MCP tool

Generates Skatteverkets begäran-om-utbetalning file (schema V6) from paid
ROT/RUT invoices — no submission API exists, the file is uploaded manually
at skatteverket.se. Headless by design for now: API routes + MCP tool
(gnubok_generate_rot_rut_file), no UI surfaces.

- lib/invoices/rot-rut-file.ts: pure XML generator with deterministic
  per-invoice blockers (hours, work type, personnummer, property info,
  mixed rot+rut, XSD limits) + 31 January deadline warnings
- rot_rut_payout_requests(+items) tables: one active begäran per invoice
  (DB triggers incl. reactivation guard), RLS, audit, pg-real tests
- Settlement: POST /settle books debit 1930 / credit 1513 via the engine
  (source_type rot_rut_payout); partial payouts → partially_paid
- Work-type lists corrected against Begaran.xsd: IT-tjänster is rut-only,
  snöskottning/tillsyn/tvätt added (schablontjänster utfört-only)
- Fix: invoice-level fastighetsbeteckning was validated but never
  persisted — now stamped onto rot lines in build-invoice-write; API
  accepts bostadsrätt pair (lägenhetsnr + BRF orgnr, editor UI deferred)
- invoice_items.brf_org_number migration + MCP scope invoices:write

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(invoices): per-company editable invoice email texts

Add an "E-posttexter" section under Settings -> Fakturering where the
subject, greeting, body and sign-off of the standard invoice email can
be customized per company in Swedish and English. Fields pre-fill with
the standard texts and only diffs from the standard are stored
(company_settings.invoice_email_texts JSONB), so future improvements to
the stock wording still reach companies that have not customized. Each
field has a reset-to-standard button; cleared fields snap back.

Texts support a fixed placeholder set (invoice number, customer name,
first name, company, due date, amount) substituted at send time in a
single pass; unknown placeholders stay literal. Custom texts are
HTML-escaped after substitution, newlines become <br> in the HTML
variant, and subject lines are flattened to a single header line.
Overrides apply to standard invoices only - credit notes, proforma and
delivery notes keep the stock texts. All send paths (UI, v1 API, MCP
approval, recurring) pick the texts up via the existing settings row.

The Zod schema half of this change (InvoiceEmailTextsSchema in
lib/api/schemas.ts) was inadvertently included in 8291f745.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(documents): accept PDFs with preamble before %PDF- header, surface content rejections as 400

detectFileMagic required the %PDF- signature at byte 0 (BOM aside),
rejecting genuine PDFs that carry a leading newline or junk bytes —
files every ISO 32000 reader opens fine. Now scan the first 1024 bytes
for the signature, matching real-reader behavior. Image types stay
strict at offset 0 to keep the anti-placeholder defense tight.

Magic-byte rejections were also mislabeled as DOC_UPLOAD_STORAGE_FAILED
(500 'Filen kunde inte sparas'), blaming storage for a client-side file
problem. Both upload routes now map them to a new
DOC_UPLOAD_INVALID_CONTENT (400) with an accurate message.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* feat(bookkeeping): full keyboard flow for manual journal entry

Enter now drives the whole verifikat flow: verifikationstext drops into
the first row missing an account, konto commits advance to debet, Enter
on an empty debet hops to kredit, and an entered amount jumps to the
next row. Once the voucher balances, Enter opens the review (unchanged
gate) and the auto-focused confirm posts it — including through the
no-underlag warning dialog. Escape in the inline review goes back to
the form.

Also fixes an Enter footgun in AccountCombobox: a bare Enter on a
freshly focused field no longer selects the first account in the list —
selection now requires typing or arrow navigation; otherwise Enter
re-commits the current value or bubbles to the form-level handler.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat: add custom inbound domains management for companies

- Implemented functionality to allow companies to claim and manage their own inbound email domains via Resend's API.
- Created a new table `company_inbound_domains` to store domain information, including status and DNS records.
- Added necessary RLS policies to restrict access based on user roles (owner/admin).
- Developed functions for domain normalization, validation, claiming, verification, and removal.
- Implemented webhook handling for domain status updates from Resend.
- Added comprehensive tests for RLS, constraints, and triggers related to the new domain management feature.

* fix: address PR #878 review findings and CI failures

- migrations: drop the ai_usage_tracking policy block from the role-gate
  migration — the table was removed by 20260504120000_remove_ai_subsystem
  and only lingers on staging as drift; a from-scratch chain (pg-real,
  Supabase preview) failed on it
- invoice-inbox: never flip a custom domain to verified off a domain.updated
  webhook alone — confirm the receiving capability with Resend first
  (fail-closed); normalize both sides of the orphan-adoption domain match
- rot/rut: block files where begärt belopp exceeds what the buyer paid
  (DEDUCTION_EXCEEDS_PAYMENT); tighten brf_org_number validation to real
  orgnr shapes; parameterize the settlement bank account (19xx, default 1930)
- rot/rut routes: log acting user on financial mutations, stop swallowing
  item mirror errors, narrow response projections (no customer ids through
  the invoice join); document the deliberate inline-XML decision
- documents: stop echoing raw storage-layer error messages to clients

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: round-2 CI + compliance findings on PR #878

- migrations: the role-gate migration targeted automation_webhooks, which
  20260515170000_webhooks_v2 renamed to webhooks on the canonical chain
  (staging kept the old name — drift); gate public.webhooks instead,
  dropping legacy schema-sync policy names defensively. Restore the
  20260623130000 owner fallback in next_voucher_number that the stale
  copied-verbatim body silently reverted (caught by engine.pg locally).
  Full migration chain verified from scratch against supabase/postgres:15.
- mcp: bump the tools/list payload ceiling 44K -> 45K — main's #877
  qualified-identifier schemas plus this branch's rot/rut tool crossed the
  ceiling only in combination; documented in the test's history log.
- rot/rut: refuse partial settlement before Skatteverkets beslut is
  recorded (would bypass the PATCH lifecycle and strand the request);
  block zero-kronor ärenden (ZERO_DEDUCTION); require sekelsiffra 16 on
  12-digit brf orgnr in both schema validation and normalizeBrfOrgNr

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: rename branch migrations off main's colliding versions

After the merge with main, two versions were shared by two files each
(20260702100000: rot_rut_payout_requests vs company_settings_dimensions_
enabled; 20260702130000: invoice_email_texts vs pending_operations_add_
create_dimension_value). psql-based CI applies by filename and doesn't
care, but Supabase branching records migrations by version (PK) — the
second file with the same version breaks the preview with a
schema_migrations_pkey duplicate. Neither branch migration is version-
recorded on staging or prod, so renaming to fresh 20260703 versions is
safe; nothing between the old and new positions depends on these objects.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(security): scope the /api MFA-gate bypass to real Bearer-auth surfaces

Any Authorization header — attacker-controlled — used to skip the AAL2
gate for every /api route, so a stolen-password AAL1 cookie session could
reach cookie-authenticated routes (which ignore the header) by attaching
`Authorization: x`. The skip is now scoped to the surfaces whose auth
contract IS the header (/api/v1 API keys, the MCP endpoint's OAuth
tokens); pure Bearer callers elsewhere (cron secret, signed webhooks)
carry no cookie session and were never touched by the gate, which only
fires for cookie users. Superagent P2 on PR #878.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* test: normalize path separators in dimension statutory guard scan

The route scan compared walked file paths against a POSIX-path allowlist,
so the suite failed on Windows (backslash separators) while passing on
Linux CI. Normalize the scanned paths to forward slashes.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 13:57:59 +02:00
Jakob Wennberg 678f2ccffd feat(mcp): P2 hygiene — honest category suggestions, skill-reference lint, cadence copy (#882)
* feat(mcp): counterparty-tied category suggestions + no_signal (P2-1)

suggest_categories padded every transaction with a company-wide
category-frequency fallback at <=0.5 confidence — an identical four-way
spread on 20+/24 items that agents correctly reported as pure noise
(agent.feedback). Real signal came from memory atoms and query_journal.

- History is now counterparty-keyed: buildMerchantHistory groups past
  categorized transactions by normalized merchant; the engine only
  surfaces history for THIS transaction's merchant, with provenance
  ('Bokförd N gånger tidigare för denna motpart') and occurrence-scaled
  confidence (0.56 at 1x, capped 0.85). No global padding — an empty
  list is the honest answer.
- The MCP tool returns no_signal_transaction_ids for transactions where
  NO source matched, steering agents to investigate (query_journal)
  instead of pattern-matching on unrelated rows.
- Both callers (REST suggest-categories route + MCP tool) share the new
  helpers, so web UI and agents improve together.

Part of dev_docs/mcp_optimization_plan.md (P2-1).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(skills): dangling-reference validation in skills:check + fix 10 dangling links (P2-2)

skills:generate/check now fail when an atom SKILL.md links a
references/*.md that does not exist on disk — a dangling pointer ships
a 404 to every agent that follows it (the weekly-booking-check
incident, agent.feedback).

The validator immediately caught 10 live dangling links in 4 atoms,
three distinct flavors:
- filename typo: swedish-asset-accounting/references/depreciaton.md
  renamed to depreciation.md (the link was right, the file misspelled)
- link mismatch: swedish-e-invoicing linked market-providers-pricing.md;
  the file is market-provider-pricing.md (link fixed)
- unauthored plans: single-shareholder-ab-fmb TODOs and reklambyra's
  'planerad utbyggnad' section used resolvable references/ paths for
  files that were never written — rephrased as plans without paths

Seed migration regenerated (4 atoms bumped, renamed reference child).

Part of dev_docs/mcp_optimization_plan.md (P2-2).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(events): align agent-feedback review cadence copy (P2-4)

gnubok_feedback replies 'we aggregate signal weekly'; the event-log
handler comment said quarterly. One of them was lying — weekly wins
(the mcp_optimization_plan triage is the living example).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 11:48:29 +02:00
Jakob Wennberg b27f6cdb04 fix(db): NULL-safe tenant guards via caller_is_company_member + mechanical sweep (#881)
The house guard 'p_company_id NOT IN (SELECT public.user_company_ids())'
skips the deny branch on UNKNOWN (NULL on either side). Not exploitable
today (company_members.company_id is NOT NULL) but a NULL p_company_id
passes the guard, and the shape fails silently under change. 9 live
functions carried it (link_*_to_voucher, reserve/release_voucher_range,
mark_entry_as_opening_balance, retag_line_dimensions,
ensure_company_dimensions, company_has_capability, rotate_company_inbox).

- caller_is_company_member(uuid): NULL-safe membership predicate
  (NULL -> false, always).
- Mechanical rewrite: every public function carrying the raw pattern is
  re-created via pg_get_functiondef with the guard swapped — deliberate
  over hand-copying 9 bodies (the stale-copy hazard behind the 07-03
  constraint clobber). Probe-validated locally: pattern swapped, NULL
  denied.
- pg-real ratchet: after full replay no public function may contain the
  raw pattern (also blocks future reintroduction); detector self-test;
  helper semantics (member/foreigner/NULL).

Existing tenant-guard suites re-assert deny semantics on the rewritten
functions in CI.

Part of dev_docs/mcp_optimization_plan.md (P2-3 follow-up, PR #872 review).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 11:48:09 +02:00
Jakob Wennberg ea236cbcdf fix(reconciliation): Bankavstämning phase 0 — correctness + feedback batch (+ nav IA regrouping) (#879)
* feat(nav): interaction-mode sidebar grouping — Arbeta/Analys/Data/Skatt & bokslut

Nav IA redesign phase 0 (dev_docs/nav_ia_redesign.md): same routes,
regrouped by what the user is doing. CLAUDE.md restructured around Hard
Rules (doc references updated); pending-page explainer removed.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(reconciliation): correctness + feedback batch for Bankavstämning (phase 0)

Engine: fetchAllRows pagination on status/run/RPC fetches (silent 1000-row
cap corrupted totals), optimistic-lock guards on manualLink + apply,
unlink audit rows attributed to the acting user (was: company UUID),
selected_matches partial apply intersected with a fresh match run.

View: silent in-place refresh instead of a full-page skeleton per action,
checkbox-gated apply with confidence badges (fuzzy unticked) in chunks of
500, honest result toasts, dry-run errors surfaced, ranked per-row picker
candidates pinned to the applied date window, currency-correct amounts
(bank side in account currency, GL side SEK), voucher links, translated
source types, colored differens, dirty-date-filter guard.

Discovery: year-end preflight 404 href fixed (/reconciliation/bank never
existed), ⌘K palette entry, real links from the transactions page.

v1: status registry schema now matches the actual ReconciliationStatus
payload, errors documented as a count, false ~0.85-threshold pitfall
replaced, route test mocks the real shape.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 11:21:38 +02:00
Jakob Wennberg 59ecaee650 feat(mcp): qualified identifiers in all tool output schemas (P1-2) (#877)
Agents grabbed the wrong id when list rows exposed a bare 'id' next to
qualified ids (journal_entry_id) with no type distinction, got
NOT_FOUND, and had to re-derive (agent.feedback).

Additive sweep of all 11 bare-id sites: every output identifier now
ships a fully qualified sibling (transaction_id, fact_id, atom_id,
company_id, dimension_id, dimension_value_id); the bare 'id' stays as
a deprecated alias for compatibility.

New ratchet test (qualified-ids.test.ts): a shrinking grandfathered
list carries the deprecated aliases; any NEW bare 'id' in an output
schema fails CI, every remaining alias must ship alongside its
qualified sibling, and stale grandfather entries must be pruned.
Convention documented in .claude/rules/mcp-server.md.

Full unit suite: 6,577 green.

Part of dev_docs/mcp_optimization_plan.md (P1-2).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 11:20:32 +02:00
Jakob Wennberg 21512db81a feat(mcp): unify missing-document surfaces on one predicate (#876)
* feat(mcp): unify missing-document surfaces on one predicate (P1-3)

The two MCP surfaces told different truths: the transactions tool keyed
'has underlag' on transactions.document_id while the verifikat tool
keyed on document_attachments — and neither respected the source-type
semantics, version chains, or journal_entry_no_doc_required waivers
that lib/worklist's canonical count applies. Measured on prod: 22,046
waived verifikat still listed to agents, 2,370 doc-exempt source types
listed, ~87 docs attached to transactions but never propagated to the
verifikat, 1,100 transactions flagged missing-receipt although their
verifikat HAS the underlag.

One predicate now lives in SQL — posted, needs-doc source type
(mirrors NEEDS_DOC_SOURCE_TYPES), no current-version doc, no waiver:

- verifikat_without_documents RPC v2 adopts the canonical predicate.
- New transactions_without_documents RPC: the bank-driven subset of the
  same predicate, joined through transactions.journal_entry_id — a
  strict subset of the verifikat surface by construction. Rows expose
  qualified transaction_id (P1-2 forward-compat); bare id deprecated.
- Both tools become thin RPC wrappers; descriptions state the actual
  set relationship.
- lib/worklist countVerifikatMissingDocument delegates to the RPC
  (previously three full-table pulls set-differenced client-side) —
  badge count and agent surfaces can no longer drift.
- Backfill: propagate transaction-attached docs to their verifikat
  where the attachment was never linked (open periods only; never
  steals a doc linked to another verifikat).

pg-real: fixture matrix (no-doc/with-doc/waived/stale-version/
doc-exempt-source/import), strict-subset assertion, per-source-type pin
of the SQL list against the TS constant, tenant guard.

Part of dev_docs/mcp_optimization_plan.md (P1-3).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(mcp): explicit grants restated + count-call comment (#876 review)

- Restate REVOKE/GRANT on verifikat_without_documents so the migration
  is self-contained (CREATE OR REPLACE preserves the 20260703130000
  grants — verified on prod: authenticated + service_role only).
- Comment on the p_limit:1 count call: total_count is computed over the
  full filtered set, independent of page size.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 10:56:14 +02:00
Jakob Wennberg 250cc7c450 feat(mcp): always-explicit retryable on structured errors + transient inference (P1-1) (#875)
Agents could not distinguish 'keep retrying' from 'stop, this is
broken' (agent.feedback): retryable was emitted only when a registry
entry declared true — absent otherwise, including for genuinely
transient DB/network failures whose SQLSTATE is lost when tools wrap
them as Error('Database error: ...').

- StructuredError.retryable is now a required boolean. Registry
  declaration wins; otherwise isTransientFailure() infers from Postgres
  SQLSTATEs (40001/40P01/57014/08xxx/53xxx/55P03), upstream HTTP
  statuses (408/429/5xx), and message signatures that survive wrapping
  (deadlock, serialization, statement timeout, fetch/socket failures).
- Unclassified transient failures surface as stable code
  TRANSIENT_ERROR (new registry entry, retryable: true).
- categorize_transaction accepts idempotency_key — the tool agents
  blind-retry after client-side approval-elicitation drops; the key
  makes that retry replay-safe instead of double-staging.
- Contract documented in .claude/rules/mcp-server.md. The planned
  'kind' field was dropped: the code registry already encodes it;
  retryable is the agent-actionable bit.

Every tool error already flows through the single dispatch point
(toToolError -> getStructuredError), so coverage is universal without
per-tool migration. Full unit suite: 6575 tests green.

Part of dev_docs/mcp_optimization_plan.md (P1-1).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 10:29:58 +02:00
Jakob Wennberg 001de3c844 feat(mcp): structured supplier-resolution failure with candidates in create_supplier_invoice_from_inbox (#873)
* feat(mcp): structured supplier-resolution failure with candidates in create_supplier_invoice_from_inbox

When supplier resolution failed (no match on id, org_number, or exact
name), the tool threw an opaque error — a dead end for the inbox
pipeline on small ad hoc vendors, which are most of the backlog
(agent.feedback). The error prose did mention supplier_id_override, but
gave the agent no candidate ids and no machine-readable next step.

Resolution failure now returns the staged-operation envelope with
staged:false: preview carries the extracted supplier identity and
near-miss candidates (normalized-name + org-digit matching — catches
punctuation/legal-suffix OCR variants like "Polarn o Pyret" vs
"Polarn O. Pyret AB", and formatted org numbers '556677-8899' vs
'5566778899'); next hints either retry-with-override on the best
candidate or a prefilled gnubok_create_supplier. Fuzzy scores never
auto-resolve — the agent confirms against the underlag.

Part of dev_docs/mcp_optimization_plan.md (P1-4).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(mcp): review fixes — EF org-number forms, candidate-pool truncation flag, override tenancy validation

Addresses the three substantive review-bot findings on #873:

- orgNumberKey(): canonical 10-digit key. Swedish orgnr is exactly 10
  digits; enskild firma org numbers are personnummer that appear in both
  10- and 12-digit forms — exact digit-equality missed that legitimate
  match, and >= 10 accepted non-orgnr garbage lengths.
- No silent caps: preview.candidate_pool_truncated + message note when
  the 500-supplier candidate pool is hit.
- The defaults fetch now validates supplier existence in THIS company on
  every resolution path and rejects a bad supplier_id_override with a
  clear error — the unresolved next-hint actively promotes overrides, so
  a bogus id must fail at staging, not opaquely at commit.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 10:14:16 +02:00
Jakob Wennberg 28827c2613 feat(dimensions): voucher-level retro-tagging workbench — reversal pairs hidden by default (#874)
* feat(dimensions): voucher-level retro-tagging workbench, reversal pairs hidden by default

UX rework of the BulkTagWorkbench (follow-up to #867): the verifikat is now
the unit of work, matching how users think ('that invoice belongs to project
X') and how every Swedish bookkeeping tool presents entries.

- /api/dimensions/tagging/lines returns voucher-grouped results via a
  two-step query: filters select QUALIFYING vouchers (line-level predicates
  become 'voucher has such a line' through the inner join), then the
  complete line set for each — tagging a voucher always covers the whole
  verifikat, never the filtered subset of one
- reversal pairs are EXCLUDED by default: an annulled entry and its storno
  net to zero in every dimension bucket as long as both sides carry the
  same tag, so retro-tagging them is a no-op with an asymmetry foot-gun
  attached; 'Visa annullerade' opts them back in, and the blocking
  motverifikat confirmation survives only in that view (correction rebooks
  remain taggable — only the original+storno pair is hidden)
- voucher rows: label, description, date, line count, distinct tag chips +
  'Delvis taggad' state, single total amount (no debit/credit columns);
  chevron expands to per-line rows (signed amount, per-line checkboxes)
  for the mixed case (a voucher split across projects)
- selection stays line-id based under the hood (the retag RPC is per
  line); shift-click ranges operate on vouchers; apply groups by resulting
  map and now chunks to the apply route's 500-line cap; failed vouchers
  auto-expand with their Swedish RPC errors inline
- 'endast otaggade' now means 'vouchers with at least one untagged line';
  the cap counts vouchers (default 150, max 300)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: pull counter-vouchers into the annulled view even outside filters (review)

Without this, a pair leg whose counter fell outside the date range would
show no motverifikat warning at all — one-sided tagging would slip
through silently, exactly the Srf U 14 skew the guard exists to prevent.
Also scope the line fetch explicitly through the parent company filter
(defense in depth).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 10:11:07 +02:00
Jakob Wennberg 45438d4d64 fix(mcp): SQL-side filtering and pagination for list_verifikat_without_documents (#872)
The tool applied min_amount in memory after the PostgREST .range() page:
total_count ignored the filter, and next_offset advanced by the filtered
row count while the DB page consumed 'limit' rows — consecutive pages
overlapped and full backlog coverage could not be proven (reported via
agent.feedback).

gross_amount is an aggregate over journal_entry_lines that PostgREST
cannot filter on, so filtering/counting/pagination move into a new
verifikat_without_documents RPC (SECURITY DEFINER with the PR #625
tenant-guard pattern; id tiebreak for total ordering; filter-respecting
total_count). Also indexes document_attachments.journal_entry_id, which
the anti-join and the link tools both hit.

pg-real invariants: disjoint + complete pages under min_amount,
filter-respecting totals, since filter, doc/draft exclusion, tenant
guard both ways.

Part of dev_docs/mcp_optimization_plan.md (P0-2).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 09:42:02 +02:00
Jakob Wennberg 4b5fe9ee57 fix(mcp): add link_document_to_voucher to pending_operations CHECK constraint (#871)
The gnubok_link_document_to_voucher tool shipped with its executor and
risk tier but its operation type was never added to the
pending_operations_operation_type_check constraint. Every real staging
INSERT failed with check_violation while dry_run previews were clean
(the INSERT is skipped). Reported 3 times by 2 companies via
agent.feedback; blocked Bokio attachment migration.

Adds a pg-real audit test that extracts every op type staged in
server.ts plus all OPERATION_RISK_TIERS keys and asserts each is
accepted by the constraint, so a staging tool can never again ship
without its constraint expansion (and a stale expand-types migration
can no longer silently drop a type).

Part of dev_docs/mcp_optimization_plan.md (P0-1).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-03 09:41:31 +02:00
Jakob Wennberg fb3f0a9cee feat(dimensions): PR9 cutover — cost_center/project become GENERATED columns, dual-write removed (#870)
The dual-write window ends (dev_docs/dimensions_implementation_plan.md PR9):
journal_entry_lines.cost_center/project are now GENERATED ALWAYS AS
(NULLIF(dimensions->>'1'/'6','')) STORED — divergence from the bag is
impossible by construction instead of by convention.

- migration 20260702230000: drift pre-flight (refuses cutover on
  inconsistent data; prod verified 0 drift across 593k rows), column swap
  (DROP metadata-only + one-rewrite ADD pair), and atomic redefinition of
  the two SQL writers — retag_line_dimensions (SET dimensions only) and
  bulk_book_transactions (INSERT names the bag only)
- TS writers stripped of the mirror spread: engine buildLineInserts
  (covers create/update/reversal), storno-service (reversal + correction),
  SIE import bulk insert, sandbox seed
- lineDimensionColumns() removed from dimension-resolver — nothing derives
  mirrors in TypeScript anymore; normalizeLineDimensions + the deprecated
  cost_center/project INPUT aliases stay (API contract, they normalize
  into the bag); JournalEntryLine ROW type keeps the fields (generated
  columns still SELECT)
- immutability carve-out unchanged BY DESIGN: its whole-row diff already
  subtracts dimensions/cost_center/project on both sides, which is exactly
  what makes it correct with generated columns (BEFORE-trigger NEW carries
  not-yet-recomputed mirror values)
- audited every reader (v1 journal-entries, MCP query_journal filters +
  group_by, rc-basis-gaps) — reads are untouched; no index, view, or
  constraint referenced the TEXT columns, so DROP COLUMN cascades nothing
- new pg suite: generated derivation, explicit-mirror-write rejection,
  draft-update recompute; existing retag/substrate/bulk-book suites
  updated to bag-only writes (their mirror assertions now exercise the
  generation expression)

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 23:28:42 +02:00
Jakob Wennberg 163fbd8222 feat(dimensions): PR8 salary — employees.default_dimensions, per-employee cost lines, aggregation re-key (#869)
Employees carry a default dimensions bag and the salary booking puts each
employee's cost on their kostnadsställe/projekt
(dev_docs/dimensions_implementation_plan.md PR8):

- employees.default_dimensions (migration 20260702220000; jsonb DEFAULT
  '{}' + object CHECK)
- salary-entries: the one-line-per-account aggregation is re-keyed to
  account+bag — P&L cost lines (löner incl. line items + base remainder,
  arbetsgivaravgifter, semesteravsättning + dess avgifter, pension, SLP)
  split per employee bag while every balance-sheet/settlement leg (2710,
  1930, 2731, 29xx, 2740, 2514) stays aggregated; liability credits equal
  the sum of the rounded debit buckets so entries balance by construction;
  dimension-less runs book byte-identically to before. Replaces the dead
  SalaryRunEmployee.cost_center/project pair (never wired)
- both book routes (dashboard + v1) read the bag via the employees join —
  read-at-book, so the run review shows exactly what will book
- employee form (new + edit) gets a gated Kostnadsställe/Projekt card;
  run review shows per-employee dims chips; run GET + v1 employee
  routes/schemas + MCP list_employees carry the field
- pre-merge audit: all salary reports (salary-journal, AGI,
  avgifter-basis, vacation-liability) read salary_run_employees — not
  journal lines — and every ledger consumer sums per account, so the
  line split breaks nothing; SIE export + dimension P&L pick the split
  up as intended

8 new engine propagation tests + book-route dims flow test.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 22:21:46 +02:00
Jakob Wennberg 755e0f7e47 feat(dimensions): PR7 producers — auto-tagged documents (invoices, supplier invoices, bulk-book, templates, MCP) (#868)
* feat(dimensions): PR7 producers — invoices/supplier invoices carry dims, generators propagate, BulkBook + templates + MCP bags

Source documents now carry dimension tags and every entry generator
propagates them onto journal lines (dev_docs/dimensions_implementation_plan.md PR7):

- invoices/supplier_invoices.default_dimensions + per-item dimensions
  (migration 20260702200000; jsonb DEFAULT '{}' + object CHECK)
- invoice-entries: issuance/payment/cash/credit propagate — item bags merge
  over the invoice default per revenue line (account+bag aggregation
  identity), payment vouchers re-propagate the linked invoice's bag onto
  every leg incl. FX result lines; ROT/RUT 1513 carries the item bag
- supplier-invoice-entries: registration/payment/cash/privately-paid/credit
  propagate with the same merge rules (expense buckets keyed account+bag)
- bulk_book_transactions RPC persists per-line bags + derives
  cost_center/project mirrors in SQL (migration 20260702201000; malformed
  bags rejected with BULK_BOOK_INVALID_DIMENSIONS); route merges the header
  default into template/manual lines
- counterparty templates: LinePatternEntry.dimensions learned from SIE
  voucher history (kept only when every occurrence agrees), applied to
  business lines on booking; QuickReviewDialog shows a dims badge
- categorize: staged dimensions bag tags business lines only (bank/VAT
  untagged); credit/convert/inbox copy paths carry bags forward
- propose-payment/send-lines stamp the invoice default so the editable
  payment grid books what the preview shows; mark-paid override lines
  accept dimensions
- UI: InvoiceEditor + NewSupplierInvoiceForm header KS/Projekt pair with
  per-row override; BulkBookDialog header default pair (both tabs)
- MCP: default_dimensions/items[].dimensions on create_invoice +
  create_supplier_invoice_from_inbox, dimensions on categorize_transaction,
  per-line bags on bulk_book_transactions — resolve-don't-select via the
  shared registry helpers, resolutions echoed

32 new propagation unit tests + 4 pg-real tests for the RPC migration.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* test: use roundOre in new dims rounding assertions (ratchet)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix: copy dimension bag per payment line, document dimensionsBagKey normalization contract (review)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 21:21:01 +02:00
Jakob Wennberg 816b1769c8 feat(dimensions): PR6 retro-tagging — audited retag carve-out, BulkTagWorkbench, staged MCP tool (#867)
* feat(dimensions): PR6 retro-tagging — audited retag carve-out, workbench, staged MCP tool

Tier-2 retro-tagging (founder decision №1, approved 2026-07-02): posted
entries in OPEN periods can have their dimension tags changed through ONE
audited path — everything about the verifikat itself stays immutable.

Carve-out (migration 20260702170000): the line-immutability trigger gains a
single narrow branch — while the transaction-local GUC set by the RPC is
active, an UPDATE of a posted line is admitted iff every non-dimension
column is unchanged, enforced by a whole-row to_jsonb diff (any future
column is protected by construction; mirrors cost_center/project are in the
changeable set because they are derived views of dimensions['1']/['6']).
Precedent: mark_entry_as_opening_balance (20260613120000).

retag_line_dimensions RPC: tenant guard (20260619130100 pattern), writer
gate (viewers rejected), posted-only, open period + company lock date
enforced, every code validated against the ACTIVE registry, immutable
dimension_retag_log row (before/after/actor/reason, INSERT-only via its own
trigger, no FKs so the trail survives hard-deletes) written BEFORE the
carve-out UPDATE. Idempotent no-op without a log row. Untag ({}) supported.
Legal position per the plan: dimensions are internredovisning metadata, not
BFL 5 kap 7§ verifikat content — this is strictly more conservative than
Fortnox/Visma (dimension-only diffs, open periods only, immutable log,
storno past locks — Tier 3 has no exceptions).

Mandatory pg suite (11 tests): GUC-less updates still blocked; amounts/
description can never change even under the GUC (transaction-local);
closed/locked/lock-date, role, registry, draft and cross-tenant rejections;
log immutability; gnubok.allow_delete bulk path unaffected.

UX (all writes through the ONE RPC): pencil on posted-voucher lines in
bookkeeping/[id] ("Påverkar endast internredovisningen, inte verifikatet")
+ retag-history card; BulkTagWorkbench at /dimensions/tagging (filters,
shift-select, merge vs "Ersätt tagg" replace mode, reversal-pair warning
with "Inkludera motverifikat" auto-selection, per-line failure display).

MCP: gnubok_tag_journal_lines (bookkeeping:write) — filter block resolved
via resolve-don't-select, ≤500 lines, staged via pending_operations (new
op type migration 20260702171000, medium risk tier, shared Zod validation
boundary between staging and commit; executor loops the RPC per line with
partial-success aggregation).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): address #867 review — SQLSTATE classification, blocking storno confirm, documented divergence

- Retag route classifies RPC errors by SQLSTATE instead of message-regex:
  P0001 (every rule violation in the RPC) → 409 verbatim, 42501 (tenant
  guard) → 403, anything else → logged 500 with a generic message. No more
  substring sniffing.
- The workbench's storno-pair warning escalates to a BLOCKING confirmation
  naming the unselected counter-vouchers before apply (Srf U 14 gross
  reporting — one-legged retags silently skew project P&L; the banner alone
  was advisory).
- The empty-bag divergence is now documented on both schemas as intentional:
  the direct dialog/workbench path allows {} (human untags phantom codes,
  logged with reason), the MCP staged path rejects it (agents never
  bulk-clear history).

Triage notes: the log's missing FKs are the point (behandlingshistorik must
survive undo_sie_import hard-deletes — a cascade would erase the trail);
SIE exports are generated fresh on demand, never cached, so post-retag
exports carry the new object lists automatically; date-scoped registry
values are deliberately not enforced at retag because entry creation does
not enforce them either — enforcing in one path only would be incoherent
(both belong to the PR10 rules engine).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 17:02:34 +02:00
Jakob Wennberg fb3fe82a56 feat(dimensions): PR5 SIE round-trip — lossless dimension import + undo lockstep (#866)
* feat(dimensions): PR5 SIE round-trip — lossless dimension import, registry upsert, undo lockstep

SIE import previously parsed and silently DISCARDED all dimension data
(object lists at sie-parser.ts:651-654, #DIM/#OBJEKT in the ignore list at
:689). Import is now lossless — the dimensions plan PR5 milestone.

Parser: #TRANS object lists ({1 "KS01" 6 "P001"}) land on the line as an
SIE-dim-no → code map (canonical numeric keys, quoted codes, malformed
pairs warn); #DIM/#UNDERDIM/#OBJEKT parse into registry records. OIB/OUB
stay ignored (dimension reporting is P&L-only in v1).

Importer (lib/import/sie-dimensions.ts): upserts missing dimensions/
dimension_values rows — never renames existing ones (ON CONFLICT DO
NOTHING); undeclared reserved numbers synthesize their SIE-standard names
(mirroring the export's orphan synthesis); codes violating the registry
CHECK are skipped with a warning but survive verbatim on lines (documented
legacy-free-text exception). Bulk voucher insert now writes the dimensions
jsonb + cost_center/project mirrors via the sanctioned dual-write helpers
(no trigger suppression needed — the immutability trigger guards
UPDATE/DELETE, not INSERT). Import auto-enables dimensions_enabled with a
result-card notice (pre-authorized by the column comment). arcim-migration
provider syncs inherit all of it via the shared parser/importer.

Undo lockstep (migration 20260702154500): created_by_import_id provenance
on both registry tables (ON DELETE SET NULL); undo_sie_import deletes the
values/dimensions the undone import introduced when no remaining
posted/reversed line references them — user-created rows and rows other
bookkeeping references are untouched. The registry guard triggers act as
backstop. replace_sie_import deliberately skips the lockstep (re-import
re-upserts the same codes). Six pg-real tests cover the lockstep.

Round-trip pinned by test: parse → import state → export → parse preserves
declarations (#UNDERDIM parent links included), values, and per-line object
lists — including synthesis of referenced-but-undeclared values.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): restate function-local statement_timeout on undo_sie_import

CREATE OR REPLACE resets proconfig, so the 290s timeout from 20260629160100
was silently dropped — regressing service-client bulk deletes to the
authenticator role's 8s limit. Caught by sie-import.replace.pg.test.ts in CI.
Full pg-real suite green (483/483, TZ=UTC).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): surface OIB/OUB drops and dimension presence as parse-level info (#866 review)

Dropping object-balance records must never be silent — one info issue counts
the skipped #OIB/#OUB rows (object-level balances are P&L-out-of-scope in
v1), and a second announces dimension data before the user executes the
import (the preview step renders parse issues), so the auto-enable notice is
no longer purely post-hoc.

Triage notes for the remaining findings: the RPC's opening SELECT is the
company-ownership check the swarm asked for; registry writes are RLS-bound;
line-verbatim codes are the documented legacy-free-text exception; export
emits no #KSUMMA so there is nothing to recompute; SIE dims 3–5 are
"reserved for future use" with no standard names, so generic synthesis is
spec-correct; ON DELETE SET NULL is deliberate — provenance is operational
metadata for undo, not räkenskapsinformation (the guarded journal lines
are), and RESTRICT would block legitimate post-retention housekeeping.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 16:05:13 +02:00
Jakob Wennberg c7f0db00c8 fix(invoices): allowlist forwarded params on the /new redirects (#861 review) (#864)
Only inbox_item_id survives the redirect to the list-page modal — the list
page reads nothing else, and arbitrary caller params must not pollute its
query string.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 15:30:21 +02:00
Jakob Wennberg 01dbef4015 feat(dimensions): PR4 reports — dimension-filtered P&L + Resultat per projekt/kostnadsställe (#862)
* feat(dimensions): PR4 reports — dimension-filtered P&L everywhere + Resultat per projekt/kostnadsställe

The Project P&L milestone of the dimensions plan (dev_docs §7 PR4).

One choke point lights up everything: generateTrialBalance gains
options.dimensions (SIE dim → code map) pushed down as jsonb containment
(dimensions @>, served by idx_jel_dimensions_gin) on both line queries, with
company-wide opening balances dropped when filtered (they cannot be
dimension-scoped; P&L-safe by whitelist). Resultatrapport, resultaträkning,
huvudbok, monthly-breakdown and the TB drill-down inherit the filter; the
KPI route filters only its P&L-side inputs (income statement, months,
expense composition) — never cash/VAT.

New report lib/reports/dimension-pnl.ts — "Resultat per projekt/
kostnadsställe" (Fortnox Resultatrapport projekt): value-as-column matrix
over one dimension with an explicit "(Utan dimension)" bucket computed as
the residual against the same trial-balance pass resultatrapport uses, so
every row and the Totalt column reconcile with the unfiltered
resultatrapport by construction. Registered in REPORT_CATALOG (visible only
when dimensions_enabled), slug-routed view + xlsx export.

UI: DimensionFilter (dimension + value picker, persistent "Filtrerad — ej
fullständig rapport" chip) mounts in FocusedReport for catalog entries
flagged dimensions: true; huvudbok rows show line dim codes.

Statutory exclusion pinned by TEST, not convention:
lib/reports/__tests__/dimension-statutory-guard.test.ts fails if the filter
parser leaks into balance sheet, balansrapport, kassaflöde, VAT, SIE or
full-archive routes/generators, or if the catalog whitelist widens.

MCP: new gnubok_get_dimension_pnl (reports:read); dimensions filter arg on
get_trial_balance/get_income_statement/get_general_ledger with
resolve-don't-select (names → registry codes, resolution echoes);
query_journal totals fixed to aggregate the FULL match set (was silently
slice-scoped while claiming otherwise) with an honest totals_scope field,
plus group_by / group_by_dimension aggregation.

Also: voucher-detail dim-6 badge now uses the registry name instead of the
non-standard "PR" abbreviation (#859 review follow-up).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): address #862 review — export disclosure, prior-column suppression, period-label honesty, route hardening

- Filtered XLSX/PDF exports now carry the partial-view disclosure past the
  file boundary (BFNAR 2013:2): filename suffix (-dim6-p001), a
  "Filtrerad … — ej fullständig rapport" row on every sheet, and a header
  note/title line in the PDFs.
- Resultatrapport drops the prior-year column when a dimension filter is
  active — project codes are time-limited under K2/K3, so "this code last
  year" may be a different project (same rule as narrowed date ranges).
- dimension-pnl no longer accepts fromDate: the matrix is cumulative from
  period_start by design (closing-balance semantics), and the period label
  now states exactly that instead of echoing a lower bound that was never
  applied. Routes/MCP tool updated to toDate-only.
- dimension-pnl routes 404 on an unknown/foreign period id and cap dim_no
  to 4 digits (matching the MCP tool's PostgREST-path guard, which the
  generator now also enforces itself).
- Statutory-guard test's generateTrialBalance call-site scan is paren-aware
  instead of a 300-char window; added fully-untagged and injection-guard
  test cases.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 15:20:47 +02:00
Jakob Wennberg 86071334cb feat(invoices): create customer & supplier invoices in modals, matching the verifikat pattern (#861)
Invoice and supplier-invoice creation now open as pop-up dialogs on their
list pages instead of navigating to standalone form pages — the same UX as
NewJournalEntryDialog (capped-height scroll, explicit-close-only so a
half-typed invoice survives stray Escape/backdrop clicks).

- InvoiceEditor gains a `bare` variant (page chrome stripped, inline actions
  replacing the fixed mobile bar, live document-type title kept) hosted by
  the new NewInvoiceDialog (sm:max-w-5xl). Draft editing pages unchanged.
- The 2,057-line supplier form moves out of the route page into
  components/supplier-invoices/NewSupplierInvoiceForm.tsx with
  bare/inboxItemId/onCreated/onCancel props, hosted by
  NewSupplierInvoiceDialog (sm:max-w-4xl).
- Modals are URL-driven (?new=1): header buttons, empty states, command
  palette, and the reports CTA all open the same dialog; browser back
  closes it. /invoices/new and /supplier-invoices/new survive as redirects
  (bookmarks, agent intents, /expenses/new alias, inbox deep links).
- The invoice-inbox "Skapa leverantörsfaktura" action opens the modal in
  place and refreshes the inbox on success instead of navigating away.

No new i18n keys; dialog titles reuse existing strings.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 14:04:52 +02:00
Jakob Wennberg 11126d6d56 feat(dimensions): PR3 tagging — voucher-form pickers, MCP dimension tools with resolve-don't-select, engine soft validation (#859)
Phase 3 of dev_docs/dimensions_implementation_plan.md. Companies with
dimensions_enabled=false see zero change; existing free-text API writers keep
working (validation is toggle-governed).

Engine (soft validation):
- validateEntryDimensions() in dimension-resolver: zero queries for untagged
  entries; toggle off → passthrough; toggle on → one settings fetch + two
  registry queries, rejects unknown dims/codes and archived values with
  Swedish per-code messages (DimensionValidationError, 400, details.issues).
  Wired into createDraftEntry + updateDraftEntry before any insert; reversal/
  storno paths untouched (verbatim copies). Fails open on transient registry
  errors — soft validation must never block bookkeeping.

MCP (agent write path):
- New tools: gnubok_list_dimensions, gnubok_list_dimension_values (fuse.js
  fuzzy), gnubok_create_dimension_value (STAGED via pending_operations —
  agents never silently mint reporting values; new op type + CHECK migration
  + executor with duplicate-idempotency).
- create_voucher/correct_entry: per-line dimensions bag + default_dimensions,
  resolve-don't-select server-side (code OR natural-language name; exact →
  fuzzy ≤0.30 with ≥0.15 runner-up margin; non-exact resolutions echoed with
  confidence; ambiguous → ranked candidates, no auto-create).
- gnubok_get_agent_briefing gains a dimensions block (enabled, dims, top
  values) — omitted when registry empty.
- TOOL_SCOPE_MAP entries; risk tier low for staged value creation.

UI:
- JournalEntryForm (manual voucher + TransactionBookingDialog embed): header
  "+ Kostnadsställe/Projekt" progressive disclosure (gäller alla rader with
  documented inheritance rule) + per-row tag popover + compact KS·PR badges;
  gated on dimensions_enabled.
- Voucher detail: display-only dimension badges with registry-name resolution.
- EditDraftEntryDialog carries line dimensions so editing a draft no longer
  strips tags.

categorize/bulk_book dims deferred to PR7 (needs the bulk_book RPC migration).

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 13:36:45 +02:00
Jakob Wennberg 6be6510d73 fix(entitlements): gate paid AI document OCR server-side (free-tier leak) (#852)
* fix(entitlements): gate paid AI document OCR server-side (free-tier leak)

Free/manual-tier companies could trigger paid Bedrock OCR (extractInvoiceFields)
with no `ai` capability check, on every transport:
- invoice-inbox HTTP paths — /upload + email /inbound (shared uploadAndExtract),
  /items/:id/attach-document, /items/:id/retry-extraction (4 call sites, zero
  capability refs);
- the gnubok_upload_document MCP tool — absent from MCP_TOOL_CAPABILITY_MAP, so a
  free-tier API key (incl. the claude.ai connector's minted gnubok_sk_ key) got
  unlimited AI extraction. This disproved the keys.ts "no MCP tool invokes AI"
  comment.

Fix (money-blocker for the free/paid tier cutover):
- Gate the 3 inbox call sites on hasCapability(CAPABILITY.ai). Upload + attach
  degrade gracefully (document still stored; extraction skipped with reason
  `no_ai_entitlement`, highest priority in the existing skipReason chain). Retry
  is an explicit "run AI now" action, so it hard-blocks with 403
  capabilityBlockedResponse.
- Register gnubok_upload_document -> CAPABILITY.ai in MCP_TOOL_CAPABILITY_MAP; the
  dispatcher already enforces the map. Correct the stale keys.ts comment and the
  misleading "deterministic field extraction" tool description + manifest copy
  (the extension migrated regex -> AI OCR).

Tests: no-AI upload/attach skip + retry 403 (sandbox-skip-extraction), retry 403
(retry-extraction), and the MCP map contract + refined dispatch<->commit parity
(capability-maps: upload_document is dispatch-only, no commit counterpart).

Self-hosted stays all-on (hasCapability short-circuits). No migration.
Follow-up (not in scope): capability-blind DashboardNav (free/paid rails identical)
and /chat gated on isVerified not `ai` — see dev_docs/nav_ia_redesign.md Part 4.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* test(mcp): assert gnubok_upload_document is ai-gated at dispatch

The gnubok_upload_document handler runs extractInvoiceFields (Bedrock OCR)
inline rather than through the entitlement-gated uploadAndExtract, so the
central MCP_TOOL_CAPABILITY_MAP dispatch check is the only paywall on that
transport. Lock it with a test (flagged by PR review as an untested money
path) so a free-tier connector key can never reach paid OCR.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-02 13:36:15 +02:00
Jakob Wennberg f675ae6565 fix(loops): committed playbook path + Sentry-first error source (#851)
* fix(loops): move playbook to committed .claude/loops.md; Sentry-first error source

dev_docs/* is gitignored ('internal reference, not published'), so the loop skills'
'read dev_docs/loops.md first' reference never resolves on main / in a fresh cloud
clone. Relocate the playbook to .claude/loops.md (committed) and repoint all 4 skills.

Rewrite loop-vercel-errors to use the wired Sentry API (the Vercel MCP is not in the
cloud routine tool allowlist) and document the required token scope (event:read,
project:read) + cloud-env secrets.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(loops): Vercel loop is Vercel-native + local (Sentry is not wired up)

Sentry is not integrated in this codebase (no @sentry/* dep, no config, no
instrumentation, zero source refs) — the SENTRY_* names in .env.local and CLAUDE.md
are leftovers. Rewrite loop-vercel-errors to source errors from Vercel (Vercel MCP
locally; Vercel API with VERCEL_TOKEN in cloud) and reclassify it as a LOCAL loop,
since the Vercel MCP is only available locally and there is no error-aggregation
service. The cloud trigger stays disabled. Only GH_TOKEN is needed to provision the
cloud loops (1 & 3).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-02 13:36:12 +02:00
Jakob Wennberg 8bb49c07a2 feat(dimensions): PR2 registry — CRUD API, register UI, settings toggle, SIE export on the new registry (#858)
* feat(dimensions): PR2 registry — CRUD API, register UI, settings toggle, SIE export on the new registry

Phase 2 of dev_docs/dimensions_implementation_plan.md. Companies with
dimensions_enabled=false (default) see zero change.

API:
- Dashboard CRUD: GET /api/dimensions (lazy-seeds system dims 1/6 via the
  ensure_company_dimensions RPC), PATCH /api/dimensions/[id] (is_system
  rename blocked), POST/PATCH/DELETE values (code immutable after creation;
  strict Fortnox code format ^[A-Za-z0-9ÅÄÖåäö_+\-]{1,20}$ at the API layer;
  retention-trigger deletes surface the Swedish "arkivera istället" message
  as 409 DIMENSION_VALUE_REFERENCED).
- POST /api/dimensions/import-existing — scans journal_entry_lines.dimensions
  for unregistered codes and mints inactive placeholder registry rows.
- v1 public API: GET dimensions + POST values (Idempotency-Key, dry-run),
  registered in the OpenAPI spec (102→104 endpoints).
- dimensions_enabled boolean on company_settings (new migration,
  UI-visibility only, never correctness-bearing) exposed through the
  existing settings read/update path.

SIE export (lib/reports/sie-export.ts):
- Reads the new dimensions/dimension_values registry; legacy
  cost_centers/projects tables now have zero readers (drop migration next).
- Fixes the latent Visma-rejection bug: #OBJEKT now declared for INACTIVE
  values referenced by lines.
- Generic-N: #DIM/#UNDERDIM loop sorted by sie_dim_no; #TRANS object lists
  serialize from the line JSONB map (sorted, '01'→'1' collapse); orphan
  codes/dims synthesize declarations from the SIE reserved-number seed —
  every referenced (dim, code) pair is guaranteed declared.

UI:
- /dimensions register (Register-recipe): tabs per dimension, search,
  sortable table, value dialog (code immutable on edit, projekt dates on
  dim 6), archive-not-delete affordances.
- DimensionCombobox shipped (mounts in the tagging PR).
- Settings toggle "Aktivera kostnadsställen & projekt" — toggle-on runs the
  import-existing scan and links to the register.
- Nav row in redovisning, rendered only when dimensions_enabled (same
  mechanism as pays_salaries).
- dimensions.* i18n namespace (51 keys, sv/en parity).
- Sandbox seed: demo dims + values, revenue line tagged {"1":"BUTIK","6":"P001"}.

Verified: 6328/6328 unit tests, guard + coverage gate green, tsc parity with
main (210=210), production build passes.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): PR2 review round — atomic archived-create, UNDERDIM ordering, import robustness, date semantics

- POST values accepts is_active so "create as archived" is atomic; the UI's
  fragile create-then-PATCH fallback is deleted (PR Agent finding 1).
- DimensionCombobox blur revert reads the committed value/values through refs
  so a selection landing inside the 150ms window always wins (finding 2).
- import-existing sanitizes candidate codes like the PR1 backfill and upserts
  with ignoreDuplicates — one bad/duplicate code can no longer abort the
  batch; created counted from returned rows (finding 3).
- SIE export emits all root #DIM before any #UNDERDIM so a parent always
  precedes a lower-numbered child (SIE4 declaration order — Swedish review);
  synthesized placeholder declarations now log one structured warning
  (BFNAR 2013:2 behandlingshistorik) + defence-in-depth comment.
- Value dates rejected (400 DIMENSION_VALUE_DATES_NOT_ALLOWED) when the
  parent dimension is flow-period (resets_annually=true); explicit null
  still clears (Swedish review).
- Sandbox seed logs seeded dimension codes; GET /api/dimensions documents
  the deliberate absence of dimensions_enabled gating (UI-visibility flag,
  not a security boundary — compliance-swarm V8.2.1 rejected by design).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 12:26:42 +02:00
Jakob Wennberg 8cc2efb083 feat(dimensions): PR1 substrate — SIE-native registry + dimensions JSONB on journal lines (#857)
* feat(dimensions): substrate — SIE-native registry + dimensions JSONB on journal lines (PR1)

Implements phase 1 of dev_docs/dimensions_implementation_plan.md:

- New company-native registry tables: dimensions (= SIE #DIM/#UNDERDIM,
  seeded is_system 1=Kostnadsställe / 6=Projekt via ensure_company_dimensions,
  nullable bare firm_id) and dimension_values (= #OBJEKT), full RLS incl.
  DELETE, audit + updated_at triggers, guard triggers (system dims undeletable,
  sie_dim_no immutable, values referenced by posted lines archive-not-delete).
- journal_entry_lines.dimensions jsonb NOT NULL DEFAULT '{}' as the single
  source of truth ({sie_dim_no: object_code}), CHECK object-typed, GIN
  (jsonb_path_ops) + partial expression indexes on dims 1/6. Inherits posted-
  line immutability from the existing trigger with zero new triggers.
- Backfill: representation copy of legacy cost_center/project text into the
  JSONB map (trigger-disabled, schema_sync precedent); legacy cost_centers/
  projects registry rows copied into dimension_values; inactive placeholder
  values for orphaned free-text codes.
- Dual-write: engine buildLineInserts + storno/correction/date-move now derive
  cost_center/project mirrors from the map via lib/bookkeeping/dimension-resolver.ts
  (normalizeLineDimensions / lineDimensionColumns); reversal copies dims.
- CreateJournalEntryLineInput + shared Zod line schema gain a dimensions bag
  (cost_center/project stay as deprecated aliases); pending-ops voucher lines
  coerce it.
- CI ratchet: direct-jel-insert check in no-new-antipatterns.mjs — inserts into
  journal_entry_lines outside sanctioned writers fail CI.
- pg-real suite: registry RLS/guards/retention, ensure_company_dimensions
  tenant guard, dims frozen on posted lines, CHECK enforcement (13 tests).

Non-breaking: companies without dimensions see zero change; no UI yet.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): address review findings — canonical keys, boundary-validated staged bags, migration guidance

- normalizeLineDimensions canonicalizes numeric keys ('01' -> '1') so
  leading-zero keys can't split values or miss the cost_center/project mirrors
  (PR Agent finding).
- New coerceDimensionsBag() in dimension-resolver is the single boundary
  validator for untyped staged payloads, enforcing the same constraints as the
  Zod line schema (string-only values, 1-40 chars, no SIE-framing chars,
  canonical keys). pending-operations normalizeVoucherLines now uses it —
  staged payloads can no longer bypass API-layer validation via numeric
  coercion (compliance-swarm V2.2/V1.2.5/PI1.1, Swedish review finding 4).
- Migration backfill comment now spells out the exact conditions under which
  the trigger-disable pattern is defensible (BFL 5:5 / BFNAR 2013:2) and what
  a future reviewer must verify before reusing it (Swedish review finding 2).
- 10 new resolver tests incl. reversal-parity (empty bag + aliases ==
  alias-only) proving the reverseEntry and storno paths normalize identically
  (PR Agent finding 1).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(dimensions): round-2 review — shared Zod schema, transactional backfill, empty-string guard

- DimensionsBagSchema now lives in dimension-resolver as the single source of
  truth; CreateJournalEntryLineSchema and coerceDimensionsBag both delegate to
  it, so the API layer and the staged pending-operations path provably cannot
  drift (compliance-swarm V2.2). coerceDimensionsBag switches to whole-bag
  semantics: any invalid entry rejects the bag, exactly like the API schema.
- Migration backfill now runs DISABLE TRIGGER / UPDATE / ENABLE TRIGGER inside
  one transaction — the ACCESS EXCLUSIVE lock from ALTER TABLE holds until
  COMMIT, so no concurrent writer can slip an unguarded line write into the
  window during a live apply (compliance-swarm V1.2, Swedish review finding 1).
- NULLIF guard: empty-string legacy mirrors can no longer mint {"n":""}
  entries the resolver would interpret as "cleared" (PR Agent round-2 edge).
- COMMENT ON dimensions.resets_annually documenting the SIE4 #IB/#OIB
  semantics the PR2+ export path must honour (Swedish review finding 2).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-02 11:27:07 +02:00
Mattsson f63d3e3100 Bug/open banking flow (#854)
* fix(enable-banking): pin Mobile BankID (decoupled) auth_method so Handelsbanken corporate connects

We never sent auth_method to Enable Banking, so it fell back to the ASPSP's
visible default — REDIRECT for Handelsbanken. For Handelsbanken *corporate*
PSUs the redirect flow does not support Mobile BankID, so authorization failed
right after the user approved in the BankID app. Mobile BankID at Handelsbanken
is a DECOUPLED method flagged hidden_method=true, which Enable Banking only uses
when requested explicitly.

Resolve the bank's preferred auth method before /auth: query the ASPSP's
auth_methods and pick the DECOUPLED (Mobile BankID) method when present,
otherwise leave auth_method unset so banks that already work are untouched.
The method name is read dynamically per psu_type, so it is robust across
sandbox/production naming.

- api-client: add approach/hidden_method to AuthMethod, fix ASPSP.auth_methods
  field name (was available_auth_methods, never populated), add
  getPreferredAuthMethod(), thread optional authMethod through startAuthorization
- index: resolve authMethod in /connect and pass it on both fresh + reconnect
- tests: cover method selection and request-body shaping

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* refactor(invoice-inbox): clean up bulk-selection toolbar UI

Redesign the selection toolbar shown when inbox items are checked:
one solid primary "Bokför valda" button with outlined secondary
actions ("Fråga assistenten", "Ta bort") and a plain selection
count. Removes the redundant "Avmarkera" button (users uncheck the
still-visible box), fixes label clipping, and gives the toolbar more
breathing room.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* chore(entitlements): bypass paywall in local development

Add isPaywallBypassed() so all gated capabilities are testable locally
without a subscription. Fires only on NODE_ENV=development (npm run dev)
or an explicit DISABLE_PAYWALL=true escape hatch — production builds run
under NODE_ENV=production and the entitlement suite runs under 'test',
so both keep exercising the real gate.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(tic): resolve enskild firma bolagsuppgifter via 12-digit personnummer

TIC's Lens search is fuzzy and only resolves an enskild firma from the 12-digit (century-prefixed) personnummer; a 10-digit form fuzzy-matched an unrelated entity. Expand personnummer to 12 digits before querying and reject hits whose registration number is unrelated to the request. Add a "Hämta" action to the settings Bolagsuppgifter panel to (re)fetch on demand.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* feat(transactions): implement categorize core for bank transaction categorization

- Added `categorize-core.ts` to handle categorization of bank transactions, supporting single and bulk operations.
- Introduced `categorizeMatchedTransaction` and `bulkBookMatchedInboxItems` functions for transaction processing.
- Implemented fiscal period validation and duplicate booking detection.
- Enhanced logging and error handling for transaction categorization.

feat(scripts): add diagnostic script for Handelsbanken ASPSP metadata

- Created `check-handelsbanken-aspsp.mjs` to fetch and display available authentication methods for Handelsbanken.
- Outputs metadata for business and personal PSU types, including default authentication methods.

fix(migrations): increase statement timeout for SIE bulk delete operations

- Updated `20260629160000_sie_bulk_delete_statement_timeout.sql` to set a longer statement timeout for bulk delete RPCs to prevent cancellations during large imports.

feat(migrations): add bulk book inbox items to pending operations

- Expanded `pending_operations` table to include `bulk_book_inbox_items` operation type in `20260630120000_pending_operations_add_bulk_book_inbox_items.sql`.
- Supports bulk booking of matched inbox items against bank transactions.

test(pg): add tests for replace_period_opening_balance_link RPC

- Implemented tests in `replace-period-opening-balance-link.pg.test.ts` to validate the functionality of the opening-balance correction flow.
- Ensured immutability of opening balance links and proper handling of posted vs. non-posted entries.

* fix(sie-export): update journal entries and lines handling in SIE export tests

* fix(migrations): resolve version collision on 20260629160000

The SIE bulk-delete statement_timeout migration shared version
20260629160000 with journal_entries_list_series_filter (merged from
main via #798/#823), causing a schema_migrations_pkey duplicate key
error on apply. Rename the branch's migration to 20260629160100.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(compliance): resolve compliance-swarm + review findings

- opening-balance/correct: compensating rollback for the non-atomic
  storno+rebook so a mid-sequence failure never leaves two posted OB
  entries (ASVS V2.3); durable audit event on every failure path
  (V16); reference the original verifikationsnummer in the corrected
  entry per BFL 5 kap 5§; document that requireWrite already enforces
  write-role + membership (V8.2.1 was a false positive)
- reports sources routes: validate the cursor date component as ISO
  (/^\d{4}-\d{2}-\d{2}$/) before use, 400 on malformed (ASVS V1.2),
  applied to both the VAT-declaration and trial-balance routes
- AgentSessionList: await the rename PATCH, revert the optimistic
  title and toast on failure (ASVS V4.5)
- bank booking: exclude same-batch siblings from the booking-time
  duplicate guard so bulk-booking distinct same-(date,amount)
  transactions no longer false-positives; pre-existing duplicate
  detection is preserved
- BulkBookInboxDialog: drop the unsafe currency-based reverse_charge
  default, add an omvänd skattskyldighet advisory, and type VAT
  options to the backend VatTreatment union
- OpeningBalanceRowEditor: hold onChange in a ref (synced in effect,
  not during render) so an unstable callback can't cause a render loop

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 18:13:00 +02:00
Jakob Wennberg 2da9c71eb3 UI badge cleanup + /chart-of-accounts route + loop skills (#850)
Bundles three separable concerns:

- style(ui): badge audit + cleanup across 45 files — real-status chips use Badge variants (raw Tailwind colors dropped), non-status count/type/label chips demoted to muted text, clustered badges consolidated; 20 unused imports removed.
- feat(bookkeeping): Kontoplan moved to a dedicated /chart-of-accounts route (nav + command palette wired); /bookkeeping shows the journal list only.
- chore(skills): loop-* automation skills + design-scan workflow under .claude/.

fix(reports): restored the destructive count badge on blocking errors in the periodisk sammanställning (EC Sales List) — a genuine status cue the audit had wrongly flattened; flagged by the PR reviewer and Swedish compliance bot, now clean.

All CI green; compliance bots report no findings.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-01 14:29:23 +02:00
Jakob Wennberg d63d2aecf0 feat: UI slop cleanup, invoice icon/header polish + year-end in Rapporter, journal-list DataList refactor (#847)
UI cleanup: removed AI-slop (redundant suppliers subtitle, decorative Sparkles glyph), decluttered the article-detail header (single status badge + muted type · #number), standardized the invoice icon Receipt→ReceiptText (no $ in a SEK app), and matched ReportExportMenu trigger size to the primary CTA on list pages.

Bookkeeping: surfaced year-end closing in Rapporter (catalog descriptor) and dropped the redundant header button; refactored JournalEntryList to DataList primitives + chunked /api/documents/counts in 50-ID batches (large pages previously 400'd); added optional fraction-digit overrides to formatCurrency. The fiscal-year lock indicator is preserved as a labeled Låst/Stängt badge in FiscalYearSelector.

All PR-bot findings triaged as false positives (unused import, formatCurrency öre, lock indicator) or intentional design (year-end placement, empty-state messaging). CI green.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 20:42:56 +02:00
Jakob Wennberg db843a7a5b fix(entitlements): gate paid MCP tools (send_invoice/agi_submit/vat_declaration_submit) server-side (#846)
The HTTP routes call requireCapability at every paid chokepoint, but the
MCP/agent path bypassed the paywall entirely: the three external-service
tools stage operations whose commit calls the email / Skatteverket services
directly, with no capability check. After the 2026-07-07 trial cutover a
trial-connected non-payer using the gnubok MCP connector could still send
invoice emails and file AGI/VAT.

Close the gap with two layers, mirroring the existing TOOL_SCOPE_MAP gate:

- Dispatch gate (mcp-server/server.ts): MCP_TOOL_CAPABILITY_MAP, checked
  right after the scope check, blocks a non-entitled company before any
  pending op is staged. Emits errorKind='capability_denied' telemetry.
- Commit-time gate (commitPendingOperation): PAID_OPERATION_CAPABILITY_MAP,
  checked before the atomic claim. The real external-service chokepoint —
  applies to the MCP approve tool AND the UI approval path, and closes the
  trial-connected-token window (the grant has expired by commit time). A
  blocked op stays 'pending', so it is re-approvable once the company subscribes.

Adds a transport-free capabilityBlockedError() helper (shared bilingual
copy) and locks both maps with tests (maps, dispatch gate, commit gate).
Only the three write/submit tools are gated; SKV read/local tools stay free
per the statutory carve-out. No DB/migration change; self-hosted stays all-on.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 16:38:14 +02:00
Jakob Wennberg 5df6199bd1 fix(invoices): remaining_amount + invoice.paid on agent mark-paid path (#825) (#845)
* fix(invoices): book remaining_amount + emit invoice.paid on agent mark-paid path (#825)

The agent/MCP commit path commitMarkInvoicePaid flipped status to 'paid' and set
paid_amount = total but never wrote remaining_amount (left at the original total)
and never emitted invoice.paid — so partial state and webhooks diverged from the
dashboard and v1 mark-paid routes.

Route the agent path through the shared planInvoicePayment helper (the source of
truth introduced in #841): compute paid/remaining/status with the overpayment
guard BEFORE booking the JE (so a rejected payment never burns a voucher number),
persist remaining_amount + the partially_paid transition, and best-effort emit
invoice.paid for webhook parity.

Adds lib/pending-operations/__tests__/mark-invoice-paid.test.ts covering the
state + event behaviour of this path.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(invoices): address review feedback on agent mark-paid path (#825)

- CAS-guard 409 message now reflects the expanded payable states: the UPDATE
  filter accepts partially_paid (reachable via a concurrent settle race), not
  just sent/overdue.
- Derive the settle amount from total − paid_amount when remaining_amount is
  null (legacy rows) instead of falling back to the full total, so a prior
  partial payment is not double-counted into a false overpayment rejection.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(invoices): derive remaining from total − paid_amount on all mark-paid surfaces (#825)

The dashboard and v1 mark-paid routes defaulted the settle amount to invoice.total
when remaining_amount was null, which over-settles a legacy invoice that has a
prior partial payment recorded in paid_amount (false overpayment / AR over-credit).
Align both with the agent path (commit.ts): remaining_amount ?? total − paid_amount.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 16:01:07 +02:00
Jakob Wennberg 46039f14f4 fix(reports): valid two-file NE-bilaga SRU submission (#318, #319) (#844)
* fix(reports): generate valid two-file NE-bilaga SRU submission (#318, #319)

The NE-bilaga "Ladda ner SRU" export produced a file Skatteverket rejects: it
was served as UTF-8 text/plain (å/ä/ö mojibake, #319) and was structurally
invalid — a single blob with #PRODUKT KONTROLLUPPGIFTER (the KU code), no
INFO.SRU/BLANKETTER.SRU split, a #SKAPAT typo, no #FIL_SLUT, and suspect field
codes 7310–7350 (#318).

Rewrite the generator to mirror the working INK2 generator: a two-file
INFO.SRU + BLANKETTER.SRU submission, ISO 8859-1 encoded and zipped, with
#PRODUKT SRU, #DATABESKRIVNING_*/#MEDIELEV_*, #BLANKETT NE-<år>P<x>,
#IDENTITET <personnummer12> <date> <time>, and #FIL_SLUT. Field codes use the
authoritative BAS NE_EJ_K1 coupling table (R1→7400 … R10→7505, R11→7440;
period dates 7011/7012). Enskild-firma identity is the owner's 12-digit
personnummer (birth-century prefix, not INK2's juridisk-person "16").

- Extract the shared ISO-8859-1 encoder to lib/reports/sru-encoding.ts (was
  inline in the INK2 route).
- Extend the NE engine/types to carry address/postort/email for INFO.SRU.
- Frontend: NE SRU download uses the INK2 blob pattern; fix a pre-existing
  param bug in EfDeclarationSection (fiscal_period_id → period_id, +format=sru).
- Add generator tests (structure, BAS field codes, zero-omission, ISO-8859-1).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(reports): address review feedback on NE-bilaga SRU generator (#318)

- getZipFilename uses the income year (fiscal year END) so the filename matches
  the blankett type/identity for broken fiscal years.
- Refuse to generate a submission when the personnummer is missing/invalid
  (compute + validate the 12-digit identity once in generateNESRUSubmission and
  throw) instead of silently emitting a placeholder #IDENTITET that Skatteverket
  would reject after upload.
- validateBlanketterSru now asserts the mandatory räkenskapsår date fields
  (#UPPGIFT 7011/7012) — their absence is a level-2 rejection.
- 10-digit personnummer century is inferred from adult age (≥18, <110) at the
  income year, fixing the e.g. 1924-born/yy=24 edge that mapped to 2024.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 16:01:02 +02:00
Jakob Wennberg f8504f3bd0 fix: audit batch — pagination truncation, MFA/dead-code cleanup, mark-paid fail-closed (#841)
* fix(reports): paginate 8 more report/ledger queries (1000-row truncation)

Raw .select() without fetchAllRows() silently caps at PostgREST's 1000-row
limit, producing wrong statutory output for high-volume companies. Following
#806 (trial-balance/VAT), wrap the remaining offenders in
fetchAllRows + a stable .order('id') + dedupeBy:

- ink2-engine / ne-engine: INK2 & NE-bilaga tax declarations under-counted
- ar-reconciliation (1510/1513), supplier-reconciliation (2440): phantom
  "Ej avstämd" gaps
- full-archive-export: 7-year DR archive (added a unique total order so rows
  are not silently skipped/duplicated across pages)
- avgifter-basis, currency-revaluation, vat-declaration

Adds a regression guard test asserting >1000 ledger lines are summed, not
truncated at 1000.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(api): close extension-dispatcher MFA gap, scope /api/events to API key, sweep dead code

Security/correctness:
- ext/[...path] dispatcher now uses requireAuth() instead of inline
  supabase.auth.getUser(), enforcing MFA (AAL2) on hosted across the whole
  enabled-extension surface (banking sync, document upload/booking, supplier
  invoices, migration). Ratchets antipatterns-baseline raw-route-auth 168->165.
- /api/events now filters by the API key's bound company_id instead of the
  user's active company (was a cross-company read with a scoped key).
- enable-banking OAuth callback calls ensureInitialized() at module load so
  the PSD2 consent audit event (ASVS V16 / GDPR Art.30) isn't dropped on a
  cold-start instance.

Dead-code sweep (all confirmed zero importers):
- delete lib/tax/calculator.ts, lib/salary/engangsskatt.ts (+test),
  lib/email/resend.ts, lib/salary/salary-transaction-matcher.ts,
  lib/webhooks/diff.ts, lib/salary/effective-values.ts,
  lib/bookkeeping/template-prompt.ts
- trim unused lib/vat/eu-countries.ts helpers (keep EU_COUNTRIES)
- remove dead getAutomaticStatus() and the abandoned Activepieces CSP entry

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(invoices): fail closed when a payment journal entry doesn't post

Three mark-paid paths (legacy route, v1 API, agent commit) diverged on the
"mark paid but the JE failed" case — two would flip the invoice to paid (or
leave an orphaned posted voucher) with no booking, silently diverging the GL
from the AR/AP sub-ledger. Unify on fail-closed:

- legacy + v1 + agent commitMarkInvoicePaid: never mark paid without a posted
  voucher; on a null/failed JE return INVOICE_PAID_BOOK_FAILED before any
  state mutation (v1 mirrors the match-invoice strict mode).
- agent path: add the .in('status',[...]).select('id') CAS guard and cancel
  the orphaned voucher (cancelOrphanedPaymentEntry) on a lost race or update
  error, matching the web route.
- legacy route: cancel the orphan on a non-race update error too (was only
  handled on the race branch).
- supplier mark-paid: stop swallowing a failed supplier_invoice_payments
  insert — that row drives the reversal amount in payment-sync; roll back the
  status flip and cancel the voucher instead.
- pending-ops orchestrator: error-check the terminal 'committed' write so an
  op stranded in 'committing' (the expire sweep only targets 'pending') is at
  least logged loudly.

Adds a guard test for the legacy fail-closed path. Full unit suite green.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(ci): unblock core build + address compliance-review findings

- avgifter-basis.ts: fix the core-build TypeScript error — PostgREST's
  type-level select parser models the salary_run embed as an array, which
  wasn't assignable to the object-typed generic. Type it `unknown` (rows are
  read via an explicit cast), making it robust across postgrest-js versions.
- /api/events: add a non-null companyId guard before the event_log query
  (defense-in-depth for the API-key-bound scope) — addresses ASVS V8.2.1 /
  ISO A.5.15.
- supplier mark-paid: add a CAS guard (.eq('status', newStatus)) to the
  payment-insert-failure rollback so a concurrent settlement can't be
  clobbered — addresses ASVS V2.3.
- dispatcher: add an AAL2 regression test asserting a non-MFA session is
  rejected (403) and the extension handler never runs — addresses the
  GDPR Art.32 review ask for the single extension chokepoint.

Verified deletions are safe: effective-values.ts was a dead duplicate — the
live AGI/payslip path inlines the same `?? override` coalescing
(generate-declaration.ts), so AGI correctness is unaffected.

next build: exit 0. Full unit suite: 6147 passing. ESLint clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 14:34:23 +02:00
Jakob Wennberg b800dcd403 style(ui): system-wide UX/UI polish pass — design-system conformance + copy cleanup (#835)
* style(ui): system-wide UX/UI polish pass — design-system conformance + copy cleanup

Multi-agent scan of all 404 UI files against the locked design system, then
141 verified surgical fixes across 109 files (net -32 lines):

- Remove forbidden elevation/motion: shadow-* and rounded-xl on cards, active:scale
  bounce, hover:shadow on list items, transition-all -> transition-colors.
- Drop font-medium from single-weight Hedvig display headings/numerals.
- Replace raw rainbow Tailwind status colors with Badge variants / brand tokens /
  neutral surfaces (achromatic chrome, semantic colors stay data-only).
- Route raw dates through formatDate(), hand-rolled currency through formatCurrency(),
  add tabular-nums to financial figures; text-gray-* -> text-foreground tokens.
- Swap hand-rolled skeletons for the Skeleton primitive; off-scale spacing -> token scale.
- Fix copy: mislabeled "Leverantörsfakturor" -> "Utgifter" on bank-import outflow total,
  collapse no-op identical-branch ternaries, broken Swedish diacritics (mojibake),
  correct mismatch-password toast, correct supplier currency-field label.
- Remove PII-leaking debug console.log on register, stray console.logs.

Verified: tsc clean on all changed files, eslint clean, production build passes.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(auth): sanitize residual error logs in register flow

Follow-up to PR review (compliance swarm V16 / GDPR Art.5(1)(f)): the
remaining console.error calls in the register flow passed raw error
objects, which Supabase may populate with PII (email) in nested fields.
Log only sanitized message strings instead.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 13:14:13 +02:00
Jakob Wennberg a8072f6423 feat: realtime updates for dashboard transactions (rebased reimplementation of #757) (#833)
* Add realtime transaction subscriptions to dashboard and transactions page

Introduce a shared browser Supabase hook so client components can create a stable realtime-capable client once and reuse it across multiple dashboard surfaces. Dashboard navigation now keeps the uncategorized transactions badge in sync through a company-scoped postgres_changes subscription on public.transactions, and the transactions page now subscribes to the same table so it can refresh its list and uncategorized count live without a full page reload.

The transactions page keeps the initial server-driven load path intact, but once mounted it listens for inserts, updates, and deletes on the active company's transactions. When a change arrives it refetches the list and total uncategorized count using the same RLS-scoped filters that already power the page, then hydrates the visible rows in the same way as the initial load. The refresh path is intentionally coalesced so bursts of realtime events do not trigger overlapping database reads.

The dashboard nav and the transactions page both use the new shared hook instead of creating browser clients ad hoc. This keeps the realtime client setup consistent, avoids duplicate client construction logic, and makes it straightforward to add more realtime dashboard consumers later without re-implementing the same Supabase plumbing.

A Supabase migration is included to add public.transactions to the supabase_realtime publication. Without that publication entry the browser subscription would be correct but silent, so the migration is required for hosted environments as well as local resets.

Also included in this commit is the current package-lock.json drift present in the staged set.

Signed-off-by: Esaias Westberg <esaias@westbergs.se>

* fix(migration): retimestamp transactions realtime publication to clear collision

20260628120000 collided with 20260628120000_ef_no_owner_employee.sql on main.
Renamed to a unique timestamp after main's latest. SQL unchanged.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* style(dashboard): fix indentation in collapsed-nav badge block

---------

Signed-off-by: Esaias Westberg <esaias@westbergs.se>
Co-authored-by: Esaias Westberg <esaias@westbergs.se>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 08:58:01 +02:00
Jakob Wennberg f8aef335c9 fix(ci): persist-credentials: false on the privileged compliance-review checkout (#831)
Stage 2 holds a write token + AWS secrets and never needs git push creds; don't
leave the token persisted in .git/config for the steps that process the
untrusted diff artifact. Closes the Superagent P2 follow-up.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-30 08:57:58 +02:00