fea5dfd1f912ec1111cec662cd8362f9aa55dfd9
340 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Jakob Wennberg
|
1d635b0d25 |
feat(receipt-hunt): look for receipts on request, from the mailbox settings page (#1496)
* feat(receipt-hunt): a button that looks for receipts on request The nightly cron exists but still does not search mailboxes, and for a good reason: a sweep of one real 172-message mailbox took over 600s, against a scheduled function's 300. Pressing a button is the honest shape for work that big. A bounded pass reports what it found and how much is left, and the person decides whether to press again; a nightly run could only truncate silently. POST /api/receipt-hunt/run searches the mailboxes for eight purchases and fetches at most ten receipts per press. Gated on the AI tier, because reading the amount out of a PDF is what makes a fetched attachment matchable at all: without it the hunt would file documents that can never pair, which is worse than not running. Writes no journal entries; every pairing is still a proposal waiting for approval. The button lives on the mailbox settings page, which already ships, and says what happened in words rather than a spinner that stops: "3 underlag hämtade. 12 köp kvar att söka igenom." huntCompany gains maxReceipts so a manual pass can carry a different budget from a nightly one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): stop a manual press spending its budget on the wrong purchases The first real press searched eight purchases, read forty mails and found nothing, which looked like a broken model. It was the ordering. Purchases are searched largest first, which is right for a nightly sweep that eventually covers everything and wrong for a button pressed a few times. On a real ledger the largest rows are the least likely to have a findable receipt: rent already invoiced, bare payment references, direct debits. Those filled the forty-mail cap, so the productive purchases further down the list, the ones whose receipts are actually sitting in the mailbox, were never read at all. The cap was the binding constraint, not the time: eight purchases and forty mails took 43s of the 300 available. A press now searches 25 purchases and reads 100 mails, measured at 85s and finding 7 underlag on the same ledger that returned 0 before. huntCompany gains maxMails alongside maxReceipts, so a manual pass can carry a different budget from a nightly one rather than sharing an environment default with it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): one underlag per purchase, and a press that fits its budget Your second press exposed three things, none of which a dry run reaches. It took 5.8 minutes. The 85s I measured was a dry run, which never fetches, uploads or extracts; each fetched receipt costs about another 37s because it is downloaded, stored, and then read by a model that opens the PDF. Seven of them ran past the 300s a serverless function gets, so in production that press would have been killed. Four receipts per press keeps a full pass inside the budget. It fetched seven receipts and proposed nothing. A single mail carries the invoice AND the receipt for one purchase under different names ("Invoice-E19DBF63-0021.pdf" beside "Receipt-2066-0204-8388.pdf"), and the same receipt reaches a second mailbox on a different message. Each was fetched separately, so the pool filled with identical candidates and the matcher refused to propose any of them rather than flip a coin. The per-run key is now the vendor and the total, which is what identifies a purchase; the filename only decides when no amount was read. Nine duplicates already in the pool were removed. And with the duplicates gone it still proposed nothing, for a separate reason: "Utlägg Norwegian" scored 0.18 against "Norwegian Air Shuttle AOC AS". Utlägg is Swedish for an expense reimbursement, bank vocabulary rather than a company, and leaving it in broke the token-subset match, so an exact 1 998 kr pair leaned entirely on a date eight days out and fell under the floor. Stripped, along with överföring, via internet, bg-bet and autogiro, in the comparison path only. normalizeMerchantName is untouched: it is the persisted konteringskarta key with a SQL mirror, and its 22 string pins and the 27-pair golden set still pass. Measured after: the Norwegian pair proposes at 0.72. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * perf(mail): read each message once per press, not once per query The third press proposed a pairing, which the second had not, but still ran 5.2 minutes against a function's 300s. Cutting receipts from seven to four had only saved 36s, which said the receipts were never the cost. Every search fetched a full message for every hit, and a press searches many purchases across every connected mailbox. One receipt mail answers several of those queries, so 25 purchases against 2 mailboxes could ask Gmail for well over a thousand messages to end up with a hundred distinct ones. Deduplication happened in the caller, too late to save the work. A mail's content never changes, so it is now read once per mailbox and kept, bounded at a thousand entries and evicting oldest first. Measured on the same ledger: 55 purchases and 100 mails now take 102s, where 25 purchases alone previously cost around 264s before a single receipt was fetched. clearMessageCache exists because tests reuse message ids and production does not. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): do not refetch a receipt the company already holds The fourth press fetched four documents the company already had: Bolagsverket, Supabase twice, Uber. They came back because I had deleted them as duplicates, and the cross-run check is the message and attachment id, which lives in the rows I removed. That was my mistake, but it exposed a real gap. The vendor-and-total key only deduplicates inside a single pass. Across passes the same purchase still arrives as an invoice in one mail and a receipt in another, with different file keys, and both were fetched: the pool fills with identical candidates and the matcher then refuses to choose between them, which is how a press can fetch four documents and propose nothing. The pass now starts from what the company already holds, so its budget goes on documents that are actually missing. Receipts per press drops to three. Measured on this ledger, a fetched receipt costs about 50s from download to a stored amount, and that is the model reading the PDF rather than the network: seven took 5.8 minutes and four took 5.1, both past the 300s a function gets. Three fits, but it is a stopgap. Doing the fetch inside the request is the wrong shape for work this slow, and the fix is to move it off the request rather than keep shaving this number. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(mail): a mail body must not outlive the run that read it Two findings from the review, both correct. The message cache held whole MailCandidate values, and one of those fields is the mail body. The contract says a body is read once to extract fields and discarded, and a process-global cache quietly broke that: bodies of one company's mail sat in memory across requests until eviction or a restart. The MailSearchService contract now has releaseCache, the Gmail adapter clears its messages, and the hunt calls it in a finally so a failed run releases them too. The duplicate key accepted an empty vendor, so two unrelated documents that happened to cost the same collapsed into one candidate. Those now fall back to the file they came from: without a vendor there is nothing to anchor an amount to. The same finding caught something worse that I had introduced one commit earlier. The persistent check derived its key from the stored extraction while the fetch derived one from the reading model, so a document filed as "Norwegian Air Shuttle AOC AS" did not recognise an incoming "Norwegian" and was fetched again. Rather than guess at aliases, which would fold "Google Cloud" into "Google Workspace", the identity is now written onto the row when the receipt is filed and read back verbatim. Rows filed before that fall back to the extraction. receiptIdentity is one exported helper with its own tests, used by both sides, instead of the same expression written twice. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): a monthly subscription is not a duplicate of last month The review caught that my duplicate key was worse than the problem it solved. Anthropic bills the same amount every month, and keying on vendor and total alone made July look like a duplicate of June: every later receipt from any recurring supplier would have been suppressed forever, silently. Duplicates block one proposal; that would have lost a receipt per month per subscription. The identity now carries the document date. Two documents for one purchase share a date; June and July do not. Two smaller faults in the same key. The amount was serialised as a raw float, so 0.1 + 0.2 read as a different total from 0.3; it is rounded to öre like every other money comparison in this codebase. And a document with no vendor was identified by its filename alone, which collapses two unrelated papers whenever a billing system attaches "invoice.pdf": those now carry the message they came from. The key is versioned so a future change to its shape cannot be mistaken for a match against rows written under the old one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
38f5d9812e |
feat(receipt-hunt): find receipts in connected mailboxes and pair them on the amount (#1492)
* feat(receipt-hunt): nightly matcher pairing unbooked purchases with held receipts Stages an attach_document_to_transaction proposal for every unbooked card purchase whose receipt the company already holds, so the underlag is attached before the transaction is booked and the gap never forms. When the user later books it, categorize-core.ts propagates the document onto the new verifikat through the matched_transaction_id link the executor writes. Deliberately scoped to UNBOOKED transactions. The posted-verifikat backlog is 96% imported history whose originals live in the previous system, so it stays a pull (the verifikat_missing_document worklist) rather than a nightly push. Ranking reuses scoreUnderlagCandidates; the pool is loaded once per company instead of per transaction, which removes both the N+1 and the newest-50 truncation a per-transaction lookup imposes on a deep backlog. Five guards, each mutation-tested: a confidence floor above the shared candidate floor, an ambiguity margin so two equally-good receipts are left to the picker rather than coin-flipped, one-receipt-one-purchase, one live proposal per purchase, and permanent suppression of pairs a human rejected. Suppression is derived from pending_operations history rather than a new table: terminal rows are immutable and a rejection is already the durable "no". Runs 05:30 UTC, after the 05:00 bank sync. Gated on RECEIPT_HUNT_COMPANY_IDS, which hunts nobody when unset so enabling it stays a deliberate act. No migration, no journal writes, no UI: proposals land in the existing Granskning queue. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(receipt-hunt): dry-run mode for provkörning against a real ledger Returns the pairings a run would stage without writing any of them, so a company can see tonight's proposals before they reach the granskningskö and so the matcher can be validated against production data without staging an operation. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(matching): fold Swedish bank descriptors so receipts reach their purchases calculateMerchantSimilarity compared raw bank descriptors, so a receipt from "Alviks kött och fisk" scored 0.125 against the bank's own row for it, "Alviks koett och fisk K3667 Kortköp/uttag" — an öre-exact pair no threshold could reach. Adds normalizeForMatch, used for similarity only, which folds what the card rails add and never changes identity: the K#### token, Kortköp/uttag verbs, a leading "Kortköp YYMMDD", trailing /YY-MM-DD dates, reference numbers glued to the name, domain wrappers, legal forms, and the three ways banks mangle Swedish letters (ö, transliterated "oe", and ?? mojibake). Processor markers become spaces because the merchant sits before the star in GOOGLE*PLAY and after it in K*IKEA GALLE. Token-subset containment is scored level with substring containment so a receipt's legal name matches the bank's trading name. normalizeMerchantName is left byte-identical and now documents why: it is a transitive input to categorization_templates.counterparty_name, a persisted UNIQUE key with a hand-written SQL mirror the ledger-context RPC recomputes at query time. Changing it would make stored keys stop equalling computed ones, so the konteringskarta join misses and insertOrUpdateTemplate inserts a second row per merchant instead of migrating the occurrence counts. Aggressive folding is safe because it is applied to both sides of every comparison, so an over-eager fold still matches; the risk is collision between different merchants, which the new tests guard. Measured on 27 receipt/transaction pairs humans actually confirmed in production: recall 27/27, and 0/7 false positives on deliberately similar but distinct merchants. Full unit suite unchanged (13,004 passing), including the 22 string pins on the frozen key path. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(mail): read-only Gmail connector so receipts are found without forwarding Forwarding was the only way a receipt reached Accounted, and it is both unpopular (97% of companies with the problem have never used their inbox address) and fragile: Arcim's own forward has been off for weeks and nobody noticed. This lets the hunt look in the mailbox instead. Scope is gmail.readonly and nothing else. It can search and download attachment bytes, and it structurally cannot send, modify or delete: the promise the consent screen makes is enforced by the grant, not by our code being careful. The consequence is deliberate: the agent can prepare a forward for a portal-link receipt but can never send one itself. Query-then-classify, never sync. For each unexplained purchase we run a provider-side search in a -3/+10 day window, pull metadata for a handful of hits, and keep nothing. No mailbox is mirrored and no message body is stored, which is what keeps this inside Google's Limited Use terms and GDPR data minimisation. Mail is searched only for purchases Underlag could not already explain, so a receipt we already hold never costs a mailbox read. The query ORs merchant against amount rather than requiring both: demanding both misses every rebrand and reseller (Anthropic bills as Claude), while the amount alone is a strong filter inside two weeks. mail_connections is service-role only with RLS enabled and zero policies, because the row holds a live refresh token and RLS cannot hide a column. Uniqueness is (company, provider, address) so a second mailbox is additive and a reconnect updates in place. Tokens are AES-256-GCM under their own key by preference, since a mail grant reads correspondence rather than backups. Core reaches the extension through a registered service, mirroring lib/email/service.ts, so lib/receipt-hunt never imports from @/extensions and a zero-extension build still compiles. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(mail): connect UI and ingest, making the hunt reach into the mailbox Two halves that together make the connector usable. Ingest (lib/receipt-hunt/ingest.ts, core): fetches the attachment, files it as a document and an inbox item with source 'mail_hunt', then stages the pairing. It lives in core because it writes documents and inbox items, and an extension may never import another extension; the mail extension only ever hands over bytes. No re-matching for a hunted receipt: it was fetched WHILE SEARCHING for a specific purchase, so the pairing is known by construction. The search is a deliberately broad OR query, which is exactly why the proposal still goes to a human with the mailbox, sender and subject written on it rather than being linked automatically. Provenance goes in channel_context, never extracted_data, because retrying extraction overwrites extracted_data wholesale and the record of which mailbox a receipt came from has to survive that. A partial unique index on (company_id, channel_context->>'mail_message_id') makes re-runs and the same receipt arriving in two mailboxes idempotent, and a 23505 is treated as success rather than an error. Guards, both mutation-tested: a duplicate message costs no provider call, and an oversized attachment is skipped rather than stored. One unreadable attachment falls through to the next and never aborts a night's hunt. UI: /settings/mail lists connected mailboxes with their health, connects a new one through a user-gesture tab (opened before the await, so popup blockers do not eat it), and disconnects behind a ConfirmDialog that states the outcome up front, including that already-approved receipts stay because they belong to the bookkeeping now. Strings in sv and en; the read-only promise is spelled out on the page rather than buried in a consent screen. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(mail): renumber migrations to clear a version collision on main 20260806150000 was already taken by preserve_preset_committed_at, and woocommerce_connections plus enforce_balance_on_posted_insert landed after this branch was cut. Two files sharing a version breaks every fresh database, which only shows up on a clean setup rather than on an already-migrated one. Applied to prod under the new versions (20260807090000 / 20260807090100), so schema_migrations matches these filenames exactly. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): make the mailbox search actually able to find an underlag A provkörning against a real ledger returned the same seven unrelated messages for every purchase, all reporting no attachments. Three separate causes, each fixed and pinned: 1. `getMessageSummary` asked Gmail for `format=metadata`, which returns headers and omits `payload.parts` entirely. Every message therefore looked attachment-free, `bodyIsReceipt` was always true, and the `found.find(c => c.attachmentIds.length > 0)` guard in the hunt could never select anything: the feature could not file a single receipt. Gmail has no format that returns MIME structure without the body, so the body now comes down the wire; it is read for nothing and stored nowhere. 2. The bank's description is not a merchant name. "Lön Juli Jakob Överföring via internet" searched for "Juli" and matched most of the mailbox. Month names and payment-rail boilerplate are now stopwords. 3. Salary and tax runs are a company's largest outgoing rows, so they consumed the whole search budget hunting receipts that cannot exist. `canHaveEmailReceipt` skips them for the mail leg only. Deliberately narrow: a supplier invoice paid over bankgiro does arrive by mail, and an "Utlägg" reimbursement has a real receipt behind it. Measured on the same ledger: 22 hits, 0 with attachments, 0 ingestable -> 4 hits, all with attachments, 3 of 4 correct (Elgiganten, Sting, Anthropic). The fourth matched a Stockholm billing address, which is why every proposal still waits for a human. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(receipt-hunt): let a model resolve merchants and pick the receipt The keyword hunt was failing for reasons regex tuning cannot reach, all measured against a real mailbox rather than assumed: - `from:anthropic.com` returns 0. Receipts arrive here by being forwarded, so the sender is the user, not the vendor. - The exact charged amount returns 0. The bank posts a converted SEK figure that appears nowhere in a USD receipt. - A date window around the purchase returns 0, while the same merchant search without one returns 10+. A forward is stamped when it was forwarded, sometimes months later. So the query now searches merchant names across the whole mailbox, and precision is restored by judgement rather than by syntax. Two model calls per run, both through forced tool use so the reply is a shape and not prose to be parsed: 1. `planMerchantGroups` resolves bank descriptors to merchants and merges repeats. Six Anthropic subscriptions become one search and one decision instead of six of each. 2. `assignReceipts` decides which mail, and which attachment on it, is the receipt for which charge, and says why in a sentence the reviewer reads. The attachment, not the message, is the unit of an underlag: a single forward routinely carries receipts for several purchases ("Fwd: Kvitton februari" has five). Migration 20260807103000 moves the dedupe key from message to message+attachment, with a backfill, because the old index would have silently blocked every receipt after the first in a forward. The model may not produce any number that reaches the ledger. It returns ids, a confidence and a reason; amounts, dates and the write stay in deterministic code. Its answer is validated, not trusted: an unknown message id, an invented filename or a low confidence drops the pairing, and any failed call proposes nothing at all. Every result still waits for a human. Measured on the same ledger: 0 receipts that could ever be filed -> 3 correct pairings (Elgiganten, Sting office invoice, Anthropic), each with a stated reason. The five remaining Anthropic charges are dated after 2026-06-15, when forwarding to the connected mailbox stopped; the model declined them correctly. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor(receipt-hunt): amount first, and drop the confidence scoring Three findings from how others build this, applied. Production email search (Superhuman, Haystack 2026) reports that recall comes from loosening retrieval and letting the model filter downstream, not from tightening the query. Retrieval depth per merchant 12 -> 25, and purchases the planner cannot name a merchant for are now searched by amount alone instead of skipped: a line like "1260525758758 Europabetalning" identifies no merchant but is a real supplier payment whose invoice may carry exactly that total. Reconciliation engines weight amount far above date (Midday: 35% vs 5%) because banks post late while amounts do not drift. The Gmail query now leads with the amount and ORs the merchant, rather than dropping the amount whenever a merchant alias exists. Still an OR: a receipt billed in USD never contains the SEK figure the bank charged. The confidence score is gone entirely. Research on verbalised confidence finds it badly calibrated, clustered on round-number anchors and barely better than chance at separating a model's own right answers from its wrong ones. That matched what this ran into: the model anchored on 0.6 / 0.7 / 0.75 / 0.9, and the 0.7 threshold discarded two correct pairings. It is replaced by an observation rather than a self-assessment, whether the charged amount is actually visible in the mail, which is what a reviewer checks first and what sorts the queue. Also fixes a real defect the run exposed: the one-file-one-purchase guard only held within a merchant group, so when the planner split one landlord into "Sting" and "Kontorsplatser" both 15 000 kr charges were assigned the same invoice. A file is now claimed once per run, which is the duplicate underlag BFL forbids. Measured on the same ledger: 3 -> 5 pairings, no duplicate. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor(receipt-hunt): harvest receipts, then pair them on the amount Splits the mailbox leg in two along the line of what each side can actually know. The model was being asked which purchase a mail belonged to. Deciding that needs the amount; the amount lives inside the PDF; a Gmail preview essentially never shows it. Measured over a real mailbox, every single pairing came back "belopp ej synligt": it was answering without the deciding evidence, which is why it declined five of six repeat subscriptions and why two correct pairings sat just under a threshold. Now it answers only what a subject, a sender and a preview line support: is this mail an underlag, and which attachment is it. Then the receipt is fetched, the extraction that already runs on document.uploaded reads its amount, date and vendor, and the pairing is the same deterministic amount-and-merchant match every other underlag goes through. Amount becomes decisive for real rather than as an instruction the model could not act on. The load-bearing fix is small: ingest now copies the extraction result onto the inbox item. The pool is read from invoice_inbox_items, so a hunted receipt with no extracted_data could never have matched anything, and the whole mail leg was quietly incapable of producing a pairing on amount. Consequences, all deliberate: - Harvesting runs BEFORE the pool is read, so a receipt found tonight is paired tonight rather than a night later. - One staging path instead of two. Mail-sourced proposals carry the same preview and confidence as every other, plus where they came from. - Deduped on the attachment filename, not on the message: the same invoice arrives as an original, a reminder and two forwards, and the old key filed "Invoice_13041840.pdf" four times over. - Capped at 8 receipts per merchant per run. Measured on the same ledger: 5 pairings attempted from thin evidence -> 16 real documents identified, each waiting on an amount it can be checked against. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * refactor(receipt-hunt): the model reads mail, arithmetic does the matching Collapses the mailbox leg to one model call that extracts fields, and hands every judgement back to deterministic code. Gone: resolving bank descriptors to merchant names, deciding which mail belongs to which charge, and the confidence score gating the result. Three prompts and two model calls become one, and mail-intelligence.ts drops from 450 lines to 250. What made this possible was measuring what a mail actually contains. The body was being downloaded and thrown away in favour of a 200-character snippet, and the body is where a forwarded receipt quotes its original sender and its original date. That is the purchase date, the thing whose absence forced the date window off entirely and made the old design miss five of six repeat subscriptions. It was there all along. So the model now answers only what text can support: is this an underlag, from whom, when, and for how much if the mail says so. Fields, not judgements. Everything after is arithmetic: - Retrieval is deterministic. No model decides what to search for. - Fetching is gated by worthFetching(): a stated amount is enough on its own, a vendor needs a plausible date, and a mail found by a purchase's own search is evidence in itself. That last rule is what handles a supplier the bank and the invoice name differently ("Kontorsplatser j BG" against "Stockholm Innovation & Growth AB"), which is what the deleted merchant-resolution call used to buy. - The pairing is the existing scorer, reached the same way as every other underlag: fetch, let the extraction that already runs on upload read the PDF, match on the amount. Amount is decisive in fact rather than as an instruction the model could not act on. Also adds the Swedish thousands-space amount formats to the query. Measured: the Sting invoice is findable as "15 000,00" and "15 000" and by no ungrouped form at all, so every amount search was missing them. Measured on the same ledger: 5 thin pairings -> 8 real documents, each with a vendor and a true purchase date, waiting on the amount in its own PDF. Currency is never converted to make a number agree. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): trust the bytes, not the mail, when filing an attachment Found by the first live run, which fetched nothing and reported success. Three defects, each invisible to a dry run because a dry run never downloads anything. 1. Gmail declares a forwarded PDF as application/octet-stream, and uploadDocument validates content against the declared type, so the upload was rejected: "Filinnehållet matchar inte den angivna filtypen". Every forwarded receipt with a generic MIME type would have failed this way, silently, since ingest swallows one bad attachment to protect the rest of the run. The type is now sniffed from the magic bytes, then the filename, and only then from what the mail claimed. 2. The filename was re-derived by a second full message fetch inside fetchAttachment, which came back empty and fell back to a generic "underlag.pdf", discarding the real "2332687551.pdf" the search had already reported. The known name now wins. 3. The provkörning script imported lib/init instead of calling ensureInitialized(), so document.uploaded reached no handler and nothing was ever extracted. It also used static imports, which are hoisted and ran before .env.local was read, leaving the extraction extension unable to build a Supabase client. Both are script defects, not product defects: the cron route calls ensureInitialized() at module level as the architecture requires. The script now loads the environment first and imports dynamically. Also makes the per-run fetch cap tunable (RECEIPT_HUNT_MAX_RECEIPTS) so a pilot can be held to a couple of documents, and adds --live to the script, which is the only way it writes anything. Verified end to end against a real ledger, every link exercised for the first time: two attachments fetched from Gmail, stored with their real names and types, extraction run on both, the amount copied onto the inbox item, and the deterministic matcher pairing Elgiganten 21 639,00 kr from the PDF against the -21 639 kr card purchase at 0.85, staged into Granskning as attach_document_to_transaction. The second document, a Bolagsverket filing receipt, carries no total and correctly paired with nothing. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(receipt-hunt): sweep a whole mailbox, and stop lending one receipt twice A backfill on a real ledger, 22 documents fetched from 172 messages. Batches the extraction (25 mails per call) so a first run on an existing company can read the whole mailbox instead of the 40 mails one call can carry, and makes the per-run caps tunable (RECEIPT_HUNT_MAX_MAILS, RECEIPT_HUNT_MAX_RECEIPTS) so a pilot can be bounded. The nightly caps stay where they are: they pace the review queue, and a backlog is a different job from a nightly tick. Two defects the backfill exposed, neither reachable from a dry run: The one-receipt-one-purchase rule only held inside a single run. `spentDocumentIds` is per-invocation, so an H&M receipt was proposed against a -358 kr purchase on one pass and a -354 kr purchase on the next, and approving both would have put the same underlag on two verifikat. A live proposal now claims its document across runs, the same way it already claimed its transaction. A document reported with no filename, on a message carrying five attachments, was not an answer but a shrug: the caller fetched attachment number one and hoped. Those are dropped now. A body-only receipt, where there is nothing to choose between, still passes. Measured after the sweep: 21 of 22 documents read correctly, and the binding constraint on this ledger is no longer retrieval but currency. Ten receipts are in SEK and five of those pair on the amount; twelve are in USD or EUR, where the bank charged a converted figure that appears nowhere in the receipt, so no comparison is possible and none is attempted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(mail): show the provider's own mark on the mailbox settings page Someone connecting a mailbox is picking an account at a provider, and the provider's mark is how they recognise which one. A generic envelope glyph said "mail" when the question is "whose". The Google "G" already existed, drawn inline inside GoogleAuthButton for the sign-in flow. It moves to components/ui/provider-marks so there is one definition rather than two, and a Microsoft square joins it for the Graph connector. Both stay inline: no external host is contacted for an icon before anyone has agreed to anything. These are the only coloured glyphs in an achromatic interface, which is deliberate rather than an oversight. A brand mark is identity, not chrome, and Google's terms require its mark unaltered rather than tinted to match a palette. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(archive): drop the duplicate mail_connections exclusion left by the rebase Main added the table to ARCHIVE_EXCLUDED_TABLES while this branch was open, so rebasing produced the key twice and the zero-extension build failed to type check. Main's entry stays, in its alphabetical place, and keeps the sentence that answers the retention question: the grants are not räkenskapsinformation, but the receipts they find are archived as documents. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(mail): record who disconnected a mailbox, without keeping the token Raised by the compliance review: disconnect() hard-deleted the row with no trace, and which mailboxes feed underlag into the books is a control over how räkenskapsinformation is produced (BFNAR 2013:2 kap 8), so switching one off should be reconstructable years later. Written by hand rather than by the write_audit_log trigger the accounting tables use. That trigger copies the whole row into audit_log, which here would mean copying an encrypted refresh token into a second table and keeping it after the entire point of the delete was to destroy it. The sibling credential table shopify_connections omits the trigger for the same reason. Only the address and provider are recorded, pinned by a test that fails if a credential ever reaches the audit entry. The review's two other flags were checked rather than assumed: nothing purges mail_hunt documents, and categorize-core.ts:403 does carry the attached document onto the verifikat when the transaction is booked. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(mail): bound every outbound call, and stop the token widening itself Four findings from the review, each checked against the code first. Neither the Gmail API nor Google's token endpoint had a deadline. Both are awaited inside Promise.all across mailboxes, so one stalled request held the whole company's hunt open until the platform killed the run. Both now carry a 15s AbortSignal, which turns a stall into one mailbox missing from tonight's sweep. `include_granted_scopes: 'true'` let Google fold scopes this app was granted elsewhere into the token issued for a mailbox, so a grant could carry more authority than the consent screen showed. Removed, and pinned by a test asserting the parameter is absent. disconnect() ignored both statement results: a failed delete still wrote an audit entry claiming the mailbox was disconnected while the credential was live, and a failed audit insert passed silently. The delete now throws, so the entry is never written for a delete that did not happen. The audit failure is logged rather than rolled back: the two can now only diverge one way, credential gone and note missing, and recreating a credential to keep them in step would be worse than a missing note. The fifth finding is real and stays open by choice, recorded in DECISIONS.md: the cron still passes searchMail=false. A sweep of one 172-message mailbox took over 600s against a maxDuration of 300, so enabling the mailbox leg nightly would time out mid-run. That flag and RECEIPT_HUNT_COMPANY_IDS get flipped together once the per-company budget is measured. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(receipt-hunt): file each attachment under its own identity Four more findings from the review. The first is a real defect. ingestMailCandidate loops over candidate.attachmentIds, but the dedupe key, the mail_attachment_id provenance and the filename were all read from index 0. Storing the second attachment therefore recorded the first one's key and name, which mislabels the row and, because the key is unique, permanently blocks the first attachment from ever landing. Masked today only because the hunt narrows to a single attachment before calling in, so nothing in the current path exercises it. All three now come from the attachment actually being stored, and the duplicate pre-check moved inside the loop so trying a second attachment is not suppressed by the first already being filed. Mutation-tested. The per-run fetch key was the bare filename, which is not an identity: "invoice.pdf" is what half the world's billing systems attach, so a second supplier's invoice would be dropped as a duplicate of the first. Scoped by vendor as well, keeping the behaviour it was written for, one fetch for an invoice that arrives as an original, a reminder and two forwards. Adds tests/pg/mail-hunt-file-dedupe.pg.test.ts for the new unique index: five attachments from one forward all land, the same attachment is refused twice, two companies hold the same file independently, other inbox sources are untouched by the partial predicate, and the message-scoped predecessor is gone. Written against CI's Postgres; there is no local DATABASE_URL here, so CI is what exercises it. --live now refuses unless RECEIPT_HUNT_CONFIRM names the same company. The script writes to whatever .env.local points at, which for this repo is production, and a recalled command should not be able to fire it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(test): cast the jsonb parameter so Postgres can type it pg-real could not determine the type of $3 inside jsonb_build_object. An explicit ::text is what the other pg tests do. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
51bf1b77bf |
feat(import): mount the knowledge-graph theater on the Arcim migrating step (#1486)
* feat(import): stream real Arcim migration progress as NDJSON (#1485) * feat(import): stream real Arcim migration progress as NDJSON The /migrate route ran the orchestrator to completion and answered with one JSON blob, so the wizard faked its progress bar: a hardcoded 55% anchor and a static step label for a phase that can take minutes. The orchestrator has had a real onProgress channel (eight emit points with Swedish step labels and anchors) since it was written; the route just never passed it. Now a request with Accept: application/x-ndjson gets a streamed response: one line per orchestrator progress event, then a terminal done line with the results or an error line carrying the same structured envelope the JSON path returns (the 200 status is already committed once the stream opens). Callers without the header keep the original single-JSON contract, so pre-deploy tabs and the existing error-mapping tests are untouched. The wizard opts in, drives MigratingStep from the real labels and anchors (mapped onto the 55-100 slice of the wizard bar), and treats a dropped connection as unconfirmed rather than failed, since the migration keeps running server-side and a blind retry could double-import. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: record the opt-in NDJSON streaming decision for /migrate Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> * feat(import): mount the knowledge-graph theater on the Arcim migrating step The migration wizard already holds the fully parsed SIE client-side (SIEData.parsed from /sie-data), so the same TheaterCanvas that carries the /import flow can build the company's knowledge graph while the migration runs: no server change, and the plain progress card stays as the fallback whenever no parsed SIE exists (e.g. providers without SIE). Unlike /import's fixed narration script, the wizard knows exactly what the server is doing: phase 1 posts one SIE file at a time and phase 2 streams the orchestrator's real progress events. ArcimMigrationTheater therefore narrates by printing those real step labels once each as they arrive, and keys the canvas to the same milestones: the GL skeleton (rings, buckets, accounts) builds during the journal writes, counterparty waves attach while customers and suppliers import, and reconciliation pulses. Real progress bar and elapsed counter stay visible throughout. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(import): review fixes on the migration theater and stream close CodeRabbit on #1486: (1) the aria-live region wrapped the per-second elapsed counter, so a screen reader re-announced the timer every second and drowned out the real step labels; the live region now covers only the narration list and the timer row is aria-hidden (the Progress bar exposes its own ARIA value). (2) controller.close() in the stream's finally block throws if the reader already cancelled, escaping start() as an unhandled rejection; now guarded like send(). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
1d01010928 |
feat(import): stream real Arcim migration progress as NDJSON (#1485)
* feat(import): stream real Arcim migration progress as NDJSON The /migrate route ran the orchestrator to completion and answered with one JSON blob, so the wizard faked its progress bar: a hardcoded 55% anchor and a static step label for a phase that can take minutes. The orchestrator has had a real onProgress channel (eight emit points with Swedish step labels and anchors) since it was written; the route just never passed it. Now a request with Accept: application/x-ndjson gets a streamed response: one line per orchestrator progress event, then a terminal done line with the results or an error line carrying the same structured envelope the JSON path returns (the 200 status is already committed once the stream opens). Callers without the header keep the original single-JSON contract, so pre-deploy tabs and the existing error-mapping tests are untouched. The wizard opts in, drives MigratingStep from the real labels and anchors (mapped onto the 55-100 slice of the wizard bar), and treats a dropped connection as unconfirmed rather than failed, since the migration keeps running server-side and a blind retry could double-import. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: record the opt-in NDJSON streaming decision for /migrate Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
ce1e0b7642 |
feat(banking): the bank-connect payoff (#1477)
* feat(banking): the bank-connect payoff: point the first sync onward Fifth activation slice. Every successful connect passes through the sync-progress dialog's done state, which used to end on a bare Klar that stranded the user on the settings panel. Now it is the payoff: the imported count stays, the work now waiting gets named (N att bokföra · M matchar fakturor, from the same worklist counts endpoint the dashboard pane refetches), and the primary action becomes Visa N att bokföra -> /transactions, where realtime rows, the match pills and the Att bokföra badge already deliver the rest. Zero-import syncs keep the plain Klar. Also mounts the fully-built-but-orphaned BankSyncSinceLastVisit pill on the transactions footer, so returning users get the same payoff line for the nightly cron (N nya transaktioner sedan sist). Strings stay hardcoded Swedish inside the extension, matching every neighboring string in the dialog. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(banking): review triage: no stale counts, no marker advance on error Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
c187fabf92 |
feat(shopify): Shopify order/refund feed into the transactions inbox (#1474)
* feat(shopify): Shopify order/refund feed into the transactions inbox
New extensions/general/shopify feed extension, modeled on the WooCommerce
feed: connect a Shopify store with Dev Dashboard custom-app client
credentials (client credentials grant, ~24h tokens, never stored), then a
nightly cron + manual sync imports paid orders and refunds via the GraphQL
Admin API (pinned 2026-07) into the transactions inbox on clearing account
1584. Feed-only: nothing auto-books. Zero PII fields are queried, keeping
the app outside Shopify's protected customer data program.
- shopify_connections migration (RLS, revoke-never-delete, encrypted
client id/secret) + shopify_sync capability and bank_sync-mirrored
backfill
- frozen external_id scheme shopify_{shop_domain}_order|refund_{id},
scoped on the shop domain so reconnects never re-import
- cursor sync on updated_at windows with 24h overlap, lock-date drop at
map time, ingest-failure cursor floor, deadline stop-and-resume,
revoked-credential flip
- /import card + settings panel, sv/en i18n, cron 03:15 in vercel.json +
regenerated Docker crontabs, logo, events, panel registry
- 65 unit tests + pg-real RLS test; extensions.schema.json enum also
gains the missing stripe entry (pre-existing drift)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(shopify): review findings from PR 1474
- token exchange: a 429 that survives every retry is throttling, not a
credential failure; stop remapping retryable 4xx to 401 so sustained
throttling can no longer flip the connection to revoked and delete the
stored credentials (CodeRabbit critical)
- order sync: advance a scanned-through watermark (run start, capped by
the failure floor) after a fully-listed window, so empty first runs and
quiet stores rotate to the back of the cron's oldest-first selection
instead of permanently occupying the 50-connection batch (CodeRabbit
major, starvation)
- add handler-level tests for the orders cron route (auth 401, disabled
503, unconfigured no-op, query failure, capability skip, happy path,
per-connection failure isolation, revoked marking)
- add 401 tests for /sync, /transaction-sync and /disconnect; pin the
cursor floor rule with a two-order page; stub the encryption key via
vi.stubEnv
- note in the panel description (sv/en) that orders can mix VAT rates and
must be split at booking (Swedish review advisory)
- DECISIONS.md: wrap underscore identifiers in backticks (MD037)
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Mattsson
|
39f4ecdad4 |
fix(providers): surface migration step errors; INK2 SRU 7104; non-modal invoice dialog (#1465)
* feat(mileage): körjournal with milersättning booking, MCP tools and CSV export New mileage_trips table (RLS, booked-delete trigger per BFL retention), lib/mileage service reusing the payroll schablon rates, /api/mileage routes (trips CRUD, period booking to 7331, salary-run push, körjournal CSV), Körjournal dashboard page + nav, and three staged MCP tools (search-only catalog). Trips book as one verifikat per period via the engine; salary path inserts mileage_taxfree line items. mileage_trips classified in the full-archive export. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor(mileage): use shared roundOre helper per tightened ratchet baseline Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): pending_operations op-type migration + Swedish review findings - New migration pair adds log_mileage_trip/book_mileage_period to the pending_operations operation_type CHECK (pg-real audit). - bookMileagePeriod refuses a period spanning several employees and names the employee in the verifikationstext when scoped (BFL motpart). - vehicle_registration required for förmånsbil trips (schema, service, MCP staging, UI surfaces the field). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): claim-first booking, CSV injection guard and driver column - bookMileagePeriod claims trips (draft to booked CAS) before creating the verifikat, so a concurrent second booking loses the race instead of double-booking; claim reverts if verifikat creation fails. - Körjournal CSV neutralizes formula-injection triggers (OWASP) and adds a Förare column naming the employee per trip. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): resolve CodeRabbit + Swedish review round: race, drift and hardening - Copying a round trip no longer re-doubles the stored distance. - pushMileageToSalaryRun claims trips before inserting line items (retry can no longer double-pay); CLAIM_LOST replaces misleading NO_TRIPS on lost races. - Booked trips are DB-immutable via a BEFORE UPDATE trigger (new migration 20260807113215): only claim/link/revert transitions and notes edits pass. - Cross-year periods rejected (schablon rates are per calendar year); payroll config year read from the date string, not TZ-dependent getFullYear(). - MCP staged bookings freeze the previewed trip set (trip_ids in params) and the commit fails on drift; validation errors return 400, not 500. - PATCH enforces the förmånsbil regnr rule on the effective row; export validates dates before they reach the Content-Disposition header; employee_id is verified company-scoped on trip creation; stale orphaned claims released. - UI: fetch flags reset in finally; ICU plural for draft summary; distance stored at the column's 1-decimal precision. - Tests: [id] route suite, pushMileageToSalaryRun suite, claim-race, drift, cross-year and update-trigger pg cases. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): revert-to-draft must clear salary_run_id at the trigger level New migration 20260807114924 replaces the booked-immutability function: a booked -> draft revert now rejects rows keeping salary_run_id, closing the DB-level double-pay path CodeRabbit flagged. pg test pins both directions; the CLAIM_LOST unit test now asserts the revert. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): company-scope employee_id on PATCH (Superagent P2) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * test(mileage): valid v4 uuid in cross-company employee PATCH test Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(providers): surface migration step errors instead of silent empty syncs A Visma company without the API module activated (403 ErrorCode 4002, "No access to module: api_standard") failed every provider call during migration, yet the wizard reported success with zero rows and mapped the 403 to "reconnect", which loops forever since OAuth succeeds against Visma's shared identity server. A real user burned time re-syncing and reconnecting, then filed the config issue as a bug. - New PROVIDER_API_MODULE_INACTIVE code; classifyProviderError reads the error body and recognizes the module error before the 403 to AUTH_EXPIRED mapping. Registry entry carries the remediation in Swedish and English (activate the API under Appar och tillagg, paid add-on on smaller plans, clear standardforetag, SIE fallback). - Orchestrator: connection-level failures (auth expired, license missing, module inactive) rethrow and abort the doomed run so /migrate answers with the typed code; other step failures stay non-fatal but land on results.stepErrors instead of only in server logs. - /preview fails fast on the two subscription codes so the user reads the remediation at connect time, before any sync. - Wizard: preview treats the new code like the Fortnox license case (CTA + SIE fallback); the result step renders error cards per cause and says "Migrering delvis genomford" instead of "Allt ar uppdaterat"; the completion toast is honest on partial failure. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(ink2): SRU field 1.1 is 7104, not 7113 (Skatteverket rejects 7113) The INK2 huvudblankett code for 1.1 Overskott av naringsverksamhet is 7104 per Skatteverket's official 2025P4 faltkoder (INK2_SKV2002-33-01-24-04). We emitted 7113, which does not exist on INK2, so filoverforing rejected every profitable company's BLANKETTER.SRU with 'UPPGIFT 7113 ar inte ett giltigt postnamn' (reported by a user for FY 2024-10-07..2025-12-31). Underskott (7114) was already correct. The wrong code originated in the swedish-sru-filing skill reference; fixed there too and regenerated the atom seed. All other emitted INK2/INK2R/INK2S codes verified against the official 2025P4 lists. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(invoices): keep the AI chat usable over the new-invoice dialog The new-invoice dialog was a modal Radix dialog: modal mode sets body pointer-events: none, aria-hidden on body siblings, and a focus trap, so the agent sheet (z-60, painted above the dialog) was visible but dead: clicks swallowed, input unfocusable, and all three dismiss paths preventDefaulted, leaving no way out except the header X. Now non-modal: page modality is restored by hand instead. A new DialogVeil primitive supplies the backdrop (Radix renders no overlay in non-modal mode) at z-40, under dialog content (z-50) and the agent sheet (z-60), and inert on #dash-shell blocks pointer, keyboard, and AT access to the page behind while the sheet (a body-level sibling) stays live. The lazy-load fallback dialog on /invoices gets the same treatment so a hung or 404'd chunk cannot dead-lock the route. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
799fa1246a |
fix(vat): downgrade per-voucher RC basis gaps only under per-rate evidence (#1464)
* fix(vat): downgrade per-voucher RC basis gaps only under per-rate evidence Per-voucher RC basis gap findings (findRcBasisGaps) blocked "Skicka till Skatteverket" as ERROR even when the flagged vouchers were legitimate moms-only rattelseverifikat whose basbelopp lives in another (often reversed) verifikat. In that state no arrangement of vouchers satisfies both the per-voucher scan and the aggregate basis/moms identity, so the block was unfixable: every correction voucher joined the blocklist it was meant to clear (Orto Engineering 3DJake support case, 2026-08). The gap finding now downgrades to a non-blocking WARNING only when ALL of the following hold, otherwise the blocking ERROR stays exactly as before: - the 44xx/45xx RC basis accounts, grouped per momssats (RC_BASIS_ACCOUNTS_BY_RATE), match ruta 30/31/32 two-sided within a 0.5 kr ore epsilon per rate; - no moms box (ruta 30/31/32) is negative; - the aggregate RC_OUTPUT_MISSING check has not fired; - the caller supplied the evidence at all (older wire payloads and totals-less contexts keep the blocking behavior). A first cross-rate-sum predicate was refuted by adversarial review: a wrong-rate fiktiv moms voucher (12% moms "covered" by a 25% basis) reached parity and unblocked a 7 800 kr under-declaration, and a net-negative rate box made the summed comparison vacuous (textbook FK004 state filing). Rutor 20-24 are partitioned by purchase type, not rate, so the certificate must come from account totals; both counterexamples plus the tolerance-hole case (shortfall inside the aggregate 0.5% tolerance still blocks) are locked in as regression tests. The evidence travels as rcBasisByRate on the declaration payload (rcBasisTotalsByRate projection), consumed by the web view and the MCP completeness checks; rc-basis-gaps.ts derives its flat account set from the same rate-grouped single source so scan and evidence cannot drift. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(vat): refuse gap downgrade on non-finite evidence; pin the ore epsilon Review findings, one pass: - CodeRabbit (major): rcBasisByRate arrives as unvalidated JSON in the web view; a missing or non-numeric field made every per-rate comparison evaluate against NaN, which compares false and PASSED the predicate, relaxing the filing gate in the unsafe direction. The predicate now refuses the downgrade outright on any non-finite basis or moms figure, covering both the web and MCP callers. - CodeRabbit (nit): added a 0.51 kr drift case so a future widening of the 0.5 kr epsilon fails a test instead of slipping through green. Declined with reasons (recorded in the PR summary): requiring textual voucher-to-voucher references before downgrading (belongs to the rattelse documentation flow, and would reintroduce the unfixable block this PR removes); epsilon stacking across rates (max 1.5 kr, immaterial at whole-krona filing and below the aggregate tolerance); explicit negative-basis guard (all negative-basis paths already block via the two-sided mismatch or the negative-moms guard, now plus the finite guard). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
4b0a185876 |
fix(invoice-inbox): parse AI extraction output wrapped in markdown fences (#1460)
* fix(invoice-inbox): parse AI extraction output wrapped in markdown fences
Since the Sonnet 5 switch (
|
||
|
Jakob Wennberg
|
70845edf69 |
feat(transactions): structured transaction_method instead of channel-in-the-name (#1459)
* feat(transactions): structured transaction_method instead of channel-in-the-name
Swedish bank feeds embed the payment channel in the description string
("Vercel Jul Överföring via internet", "ANTHROPIC* ... Kortköp/uttag"):
the PSD2 remittance array is joined into one string and the ISO 20022
type codes were dropped at insert. This promotes the channel to data:
- transactions.transaction_method (text + CHECK closed vocabulary: card,
transfer, bankgiro, plusgiro, swish, autogiro, e_invoice, international,
deposit, withdrawal, salary, fee, interest, adjustment) plus verbatim
bank_transaction_code / proprietary_bank_transaction_code evidence
columns (data_quality_master Appendix B "Layer-A capture").
- classifyTransactionMethod() in lib/transactions/transaction-method.ts:
explicit source method (Stripe txn.type) > trailing Swedish channel
phrase > ISO 20022 family/subfamily > proprietary-code keywords > MCC.
It also splits the clean display title off the description.
- Ingest stores the clean title as description and the full bank string
as original_description; dedup is untouched (external_id is date+öre,
the content bridge reads original_description and is prefix-based, and
a trailing strip leaves a prefix). Enable Banking passes the codes
through; the Stripe feed sets methods from its balance-txn types.
- Backfill migration classifies existing rows from the description text
(+ MCC and Stripe prefixes) and strips unedited titles; user-edited
titles are never rewritten.
- mapping-engine also matches original_description so user rules written
against the full bank text keep firing.
- UI: the inbox row shows the clean name; clicking it now folds out
"Betalsätt: Kortköp" etc. (sv/en), making every classified row
expandable.
A card purchase implies a physical receipt, a Bankgiro/e-invoice payment
implies a supplier invoice: downstream automations can now branch on the
rail instead of regexing display strings.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(transactions): anchor counterparty-template identity on original_description
Audit follow-up to the phrase-strip change: counterparty template lookup
AND learning derived their key from merchant_name || description. With
the working title now stripped ("SPOTIFY AB Kortköp" -> "SPOTIFY AB"),
templates learned from the full bank string would only re-match via the
occurrence-gated single-token tier, and single-token counterparties with
fewer than 3 bookings would silently stop matching.
Both sides now read merchant_name || original_description || description:
the immutable bank original is identical across eras (and across user
renames), so every stored key and alias keeps matching exactly. Same
anchoring rationale as buildMerchantHistory in category-suggestions.
Existing tests that relied on the fixture's default original_description
now state it explicitly; two new regression tests pin the era stability
(lookup via alias on the full string, learning key derivation).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(transactions): review follow-ups on method classification
- methodFromCodes: two-pass subfamily-then-family scan so a SALA/XBCT
refinement on the proprietary code beats a bare family match on the
ISO code, matching the documented precedence; pinned by a test.
- mapping-engine: regression tests for merchant/description patterns
that only match original_description, including the invalid-regex
substring fallback and the no-match default.
- Stripe: regression test for the SDK-unmodeled 'tax' balance-txn type
mapping to 'fee'.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(transactions): scope method classification to feed rows + adjective guard
Multi-bank risk hardening before the backfill ships:
- Feed-row scope: classification and title stripping now require a real
import feed (import_source present, not manual/mcp), both at the
ingest boundary (USER_CREATED_IMPORT_SOURCES, now exported) and in
every backfill statement. User-authored titles like "Egen insättning"
on manual/MCP rows are never classified and never rewritten.
- Adjective guard (TS + SQL): a strip that would leave the title ending
in a possessive/scope adjective (egen/eget/privat/intern/extern ...)
is skipped, so "Egen insättning" stays whole even on bank-feed rows;
the method column still classifies (deposit).
- Unknown bank phrasings remain untouched by construction: an unmatched
phrase means no method and no rewrite, so the worst case for any bank
whose vocabulary we have not seen is the status quo.
Pinned by new unit + pg-real cases (user-created exclusion for
NULL/manual/mcp, adjective guard, feed defaults in the pg fixture).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* chore(migrations): re-timestamp transaction_method migrations after rebase
Main gained migrations dated 20260729-20260730 (already applied to prod)
while this branch carried 20260728 versions, which would have applied
out-of-order on merge. The files have never reached prod, so renaming to
current timestamps is safe and removes any dependence on the integration's
out-of-order handling. All code/doc references updated; the pg test reads
the backfill by its new filename.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* chore(migrations): bump transaction_method versions past prod's max
Main's newest applied migration is 20260730090000 (future-leaning
timestamp), so the previous 202607300731xx rename still sorted before
prod's tail and risked a silent skip on merge-time apply. Versions are
now 20260730100000/20260730100100, strictly after everything applied to
prod. References updated; full migration stream replays clean.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(transactions): final review round: keyboard guard + bank_connection_id feed marker
- TransactionInboxCard: row-level Enter/Space handling now ignores events
bubbling from nested controls, so keyboard activation of Bokför / the
overflow menu is no longer cancelled by the (now much more common)
expandable row.
- Feed predicate parity with isImportedTransaction(): a live
bank_connection_id marks a feed row even when import_source is unset
(the oldest PSD2 rows predate that column), in both the ingest
classifier and every backfill statement: those legacy rows now get
classified instead of being skipped as user-created.
- pg fixture typing uses the TransactionMethod union.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* chore(migrations): re-timestamp transaction_method migrations past prod's 20260807 tail
Prod max applied is 20260807170000 (verified by name via list_migrations);
the 20260730-stamped pair would sort before it. References in code,
tests, and DECISIONS.md updated to the new versions.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(migrations): enforce, not assume, original_description preservation in the title strip
The strip UPDATE now fills a NULL original_description from the
pre-strip description in the same statement. Prod has zero such rows
(0/25,566 feed-scope rows, verified read-only), and 20260605120000's
backfill plus ingest make the NULL case unreachable on any DB that
replayed history, but the migration should not depend on that history
to avoid losing the only copy of a bank string.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* docs: record the compliance-review triage of the backfill's booked-row title strip
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
|
||
|
Jakob Wennberg
|
a49d75db77 |
fix(migration): Visma pagination + chunk-insert resilience (the '300 misslyckades' case) (#1455)
* fix(providers): paginate Visma eAccounting with $page/$pagesize eAccounting silently ignores OData $top/$skip, so every request returned page 1 and getPaginated appended the first page TotalNumberOfPages times: customers were imported in triplicate and invoice chunks hit unique violations. Also stop on an empty page so a stale Meta can never loop or duplicate. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(migration): survive bad rows in entity imports instead of failing whole chunks One PostgREST insert per 500-row chunk is all-or-nothing, so a single duplicate reported every row as failed ('300 misslyckades') with no cause shown. Now: dedupe repeats within the fetched data (paging faults, source duplicates), fall back to per-row inserts when a chunk is rejected, store empty invoice numbers as NULL instead of colliding '', surface the first DB error in the result UI, and mark all-failed steps with an error icon. Sales invoices also carry remaining_amount so open invoices no longer land as settled. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(migration): never per-row retry after a successful bulk insert with short read-back A succeeded statement whose .select() returns fewer rows than sent means the rows ARE in the table; retrying them one by one would duplicate every unreturned row. Pair what came back and report the tail instead. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(migration): count stub-insert casualties as failed and sample enrichment errors Review follow-ups: invoices dropped because their customer/supplier stub insert errored are DB failures, not matching misses; classifying them as noMatch rendered a green result row with the database error hidden. Enrichment failures now also feed errorSample. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
7411a0171b |
feat(mileage): körjournal with milersättning booking, MCP tools and CSV export (#1448)
* feat(mileage): körjournal with milersättning booking, MCP tools and CSV export New mileage_trips table (RLS, booked-delete trigger per BFL retention), lib/mileage service reusing the payroll schablon rates, /api/mileage routes (trips CRUD, period booking to 7331, salary-run push, körjournal CSV), Körjournal dashboard page + nav, and three staged MCP tools (search-only catalog). Trips book as one verifikat per period via the engine; salary path inserts mileage_taxfree line items. mileage_trips classified in the full-archive export. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * refactor(mileage): use shared roundOre helper per tightened ratchet baseline Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): pending_operations op-type migration + Swedish review findings - New migration pair adds log_mileage_trip/book_mileage_period to the pending_operations operation_type CHECK (pg-real audit). - bookMileagePeriod refuses a period spanning several employees and names the employee in the verifikationstext when scoped (BFL motpart). - vehicle_registration required for förmånsbil trips (schema, service, MCP staging, UI surfaces the field). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): claim-first booking, CSV injection guard and driver column - bookMileagePeriod claims trips (draft to booked CAS) before creating the verifikat, so a concurrent second booking loses the race instead of double-booking; claim reverts if verifikat creation fails. - Körjournal CSV neutralizes formula-injection triggers (OWASP) and adds a Förare column naming the employee per trip. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): resolve CodeRabbit + Swedish review round: race, drift and hardening - Copying a round trip no longer re-doubles the stored distance. - pushMileageToSalaryRun claims trips before inserting line items (retry can no longer double-pay); CLAIM_LOST replaces misleading NO_TRIPS on lost races. - Booked trips are DB-immutable via a BEFORE UPDATE trigger (new migration 20260807113215): only claim/link/revert transitions and notes edits pass. - Cross-year periods rejected (schablon rates are per calendar year); payroll config year read from the date string, not TZ-dependent getFullYear(). - MCP staged bookings freeze the previewed trip set (trip_ids in params) and the commit fails on drift; validation errors return 400, not 500. - PATCH enforces the förmånsbil regnr rule on the effective row; export validates dates before they reach the Content-Disposition header; employee_id is verified company-scoped on trip creation; stale orphaned claims released. - UI: fetch flags reset in finally; ICU plural for draft summary; distance stored at the column's 1-decimal precision. - Tests: [id] route suite, pushMileageToSalaryRun suite, claim-race, drift, cross-year and update-trigger pg cases. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): revert-to-draft must clear salary_run_id at the trigger level New migration 20260807114924 replaces the booked-immutability function: a booked -> draft revert now rejects rows keeping salary_run_id, closing the DB-level double-pay path CodeRabbit flagged. pg test pins both directions; the CLAIM_LOST unit test now asserts the revert. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mileage): company-scope employee_id on PATCH (Superagent P2) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * test(mileage): valid v4 uuid in cross-company employee PATCH test Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
93f81f03e8 |
feat(providers): WINT migration provider behind WINT_MIGRATION_ENABLED (#1446)
* feat(providers): WINT migration provider behind WINT_MIGRATION_ENABLED Adds WINT (wint.se) as a sixth migration provider, built against the OpenAPI specs WINT's own API host serves publicly. Tier A scope: only the partner-facing v1 endpoints are used; the general ledger is fetched as vouchers/accounts and rendered as SIE 4E by our own sie-builder, with opening balances for earlier years derived backward from the current-year Ib anchor. Auth is the user's WINT login exchanged once for a JWT pair; the password is never stored. Ships dark: the wizard shows a disabled "Kommer snart" card, and the server-side /connect gate rejects WINT until WINT_MIGRATION_ENABLED=true. Live verification against a real WINT account is still outstanding. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(providers): harden WINT provider per PR #1446 review findings Addresses CodeRabbit and Swedish accounting review feedback in one pass: - Ib anchor selection now uses WINT's unfiltered fiscal-year list, so an active year outside the allowed import window can never silently anchor the wrong year; the voucher chain is extended through the anchor and a per-year fetch failure fails that year loudly instead of sinking the whole migration. - Auth token exchange is strict: only LoginState Success with a complete access+refresh pair mints a consent (a pair without a refresh token is unrefreshable and would break days later). - WintApiError no longer retains full response bodies (bounded 300-char diagnostic; bodies can carry customer data and errors get logged). - sie-builder refuses to render structurally invalid vouchers (missing account number or booking date) and documents deleted-voucher gaps in a #PROSA record per BFL 5 kap 6-7 §. - Account classification: 20xx is equity, 83xx is financial income. - SIE validator accepts EUBAS97 as BAS-based (standard kontoplanstyp; it previously produced a false non-BAS warning on every WINT/Bollbok file). - New tests: resolveConsent WINT refresh flow, credential upsert payload (no mail/password persisted), WINT fetch failure path, EUBAS97 warning regression, builder invalid-data rejection, vi.clearAllMocks hygiene. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(import): pin EUBAS97 acceptance to the exact SIE spec value Review follow-up on PR #1446: match EUBAS97 exactly instead of any EUBAS* prefix, so the non-BAS kontoplan warning stays pinned to the four kontoplanstyp values the SIE 4B spec enumerates (BAS95, BAS96, EUBAS97, NE2007) rather than silently accepting unknown future variants. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
707d597b2e |
feat(woocommerce): store order/refund feed extension (#1442)
* feat(woocommerce): store order/refund feed extension Connect a WooCommerce store via the wc-auth key handshake (manual key fallback) with per-store consumer key/secret AES-256-GCM encrypted at rest, and import paid orders and refunds into the transactions inbox as a bank-style feed on the 1680 cash account. Feed-only: nothing auto-books, gateway fees/payouts are out of scope (core wc/v3 does not expose them). Sync is cursor-paginated on modified_after (offset pages only inside same-second date_modified ties), terminates on an empty page, holds the cursor below failed refund fetches / ingest errors / deadline-skipped work, checks the time budget between refund fetches, and drops rows dated on or before bookkeeping_locked_through on every run. Nightly cron gated on the extension registry + new paid capability woocommerce_sync (backfilled to existing bank_sync grant holders). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(migrations): move woocommerce migrations past main's 20260806090000 origin/main gained 20260806090000_recurring_schedule_interval_months while this branch was in flight; identical version timestamps abort the Supabase apply, so the two new migrations move to 20260806170000/20260806170100. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(woocommerce): resolve CodeRabbit review findings - callback 503s early when WOOCOMMERCE_CREDENTIALS_ENCRYPTION_KEY is unset: encryptCredential would otherwise throw after the probe and strand the pending row without error_message - disconnect and upstream-revoke clear the encrypted consumer key/secret: nothing reads them after revoke and keeping decryptable dead credentials is unnecessary retention - manual sync gets a 240s time budget and the panel reports a truncated run as 'partial, sync again' instead of a normal completion - listOrderRefunds terminates on an empty batch (hosts may cap per_page), dedupes by id against hosts that ignore page, and caps total pages - unparseable money strings count as errors and log instead of being silently identical to a zero total - pg test uses per-run unique store URLs so committed rows cannot hit the store_url partial unique index across pg-real runs Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(woocommerce): resolve CodeRabbit cycle-2 findings - listOrderRefunds throws when the page cap is exhausted with data still flowing, instead of returning a silently partial list the sync cursor would advance past; the error routes into the existing held-cursor refund-retry path - partial sync results keep the row-error count, and the partial toast string surfaces it (ICU plural, hidden at zero) in both locales Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * chore: retrigger CI after dropped push event Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
a0ca692fed |
feat(invoices): quarterly, half-yearly and yearly recurring invoice schedules (#1438)
* fix(mcp): offer the link tool in the uncategorized-transactions VAT blocker The gnubok_vat_close_check blocker hint only named categorize/auto-match, both of which create new bookkeeping. For a transaction whose affarshandelse is already booked on an existing verifikat, following the hint would double-book, so agents dead-ended the case into "contact support" (2026-08-06 support mail from Orto Engineering). The hint now also names gnubok_link_transaction_to_journal_entry, is extracted as an exported constant pinned by a test, and the tool joins the categorize_month recommended loadout. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(invoices): quarterly, half-yearly and yearly recurring schedules User request: recurring invoice schedules only supported monthly cadence. Adds interval_months (SMALLINT 1-12, default 1) to recurring_invoice_schedules; the UI offers manadsvis/kvartalsvis/ halvarsvis/arsvis presets while API and MCP accept any 1-12. The cron advances next_run_date by whole intervals from the due date, and the new rollNextRunDateForward() helper rolls missed or edited interval schedules on their own month grid so a quarterly Jan/Apr/Jul/Oct schedule missed in an outage rolls Jan 15 to Apr 15, never Feb 15. Monthly (interval 1) keeps its existing today-anchored recompute semantics unchanged. Changing the interval alone never touches next_run_date: the new cadence applies from the next run, so an edit can never pull a send earlier. Existing rows default to 1 and behave byte-identically. The MCP slice of this feature (interval_months on the three recurring-schedule tools in server.ts) was committed in d2600907f alongside the VAT-blocker hint fix by a parallel session sharing this worktree. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(invoices): address PR #1438 review findings CodeRabbit round 1, all three findings: - MCP descriptions now state the full accepted interval range (any integer 1-12) instead of enumerating only the 1/3/6/12 presets, and qualify that changing ONLY interval_months leaves next_run_date untouched. - assertValidCadence rejects fractional day_of_month. - rollNextRunDateForward rejects calendar-invalid anchors that pass the shape regex (2026-13-05, 2026-02-31), with regression tests. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
d41ef2a909 |
feat(sandbox): seed payroll, articles and a year of ledger history; calm the connect CTAs (#1437)
* feat(sandbox): seed payroll, articles and a year of ledger history; calm the connect CTAs The sandbox showed neither Löner nor a usable set of reports, and the "connect X" surfaces were oversized boxed cards. Sandbox seed: - pays_salaries + employer_registered, so Löner and Anställda appear at all (an enskild firma is not an employer by default). Both seeded employees are employment_type 'employee': an EF may employ staff, just not its own owner. - Two employees, one booked and one open lönekörning, and the three verifikat the booked run must have posted (7210/2710/1930, 7510/2731, 7290+7519/ 2920+2940). Skatteavdrag comes from the real Skatteverket 2026 tables. - Year-to-date ledger history, January through last month, with the quarterly momsredovisning cleared to 2650 and paid on the SFL deadline. Without the settlement the demo collected VAT all year and never remitted it, which left an implausible bank balance and 155 813 kr of moms "att betala". - The history is exempted through journal_entry_no_doc_required, the same way the SIE-import opt-in treats imported books: its kvitton live in the previous system, and unflagged it put 39 "verifikat utan underlag" on the home screen. - Artikelregister, and the BAS accounts the K1 chart omits for an enskild firma. - History is numbered before the invoice and payroll vouchers so the series runs forwards through the year, and its writes are batched. Connect CTAs: - Bank picker: a two-column grid of 95px bordered logo cards becomes flat hairline rows, Lucide icons, and a quiet inline connecting state. - Cloud backup: each provider collapses to one row; the BFL note is shown once for the section and names only configured destinations. - Hem first-run: only the active step argues its case, but every not-done step keeps a reachable action. The Skatteverket nudge becomes one quiet sentence. Mobile assistant FAB: a fresh open is desktop-only, since the bottom nav already has an Assistent tab. A collapsed session keeps its handle everywhere except /chat, which is itself the way back to the conversation. Also closes a real hole: /api/salary/runs/[id]/payslips/send had no sandbox guard, and a seeded booked run put "Skicka lönebesked" one click from an anonymous visitor with live Resend behind it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(sandbox): check the two unchecked Supabase errors and tighten review nits CodeRabbit review on #1437. Major: two calls discarded their error and continued with null data. A failed chart_of_accounts re-select would have written account_id: null onto every ledger-history and salary voucher line, and a failed next_voucher_number would have inserted a posted verifikat with no number, which is a hole in the verifikationsserie (BFNAR 2013:2). Both now throw, and a null voucher number is rejected explicitly. Minor: the A-004 note claimed a 10 % markup on numbers that are 11.1 %; the salary breakdown test's name said the opposite of its assertions after the switch to the real tax table; the ledger-history doc still said 4 to 6 verifikat per month before the quarterly momsredovisning added a seventh in March, May and June. Bank picker: the spinner is aria-hidden, so loading and connecting had no text equivalent and a failed bank fetch was never announced. Added role="status" with an sr-only label, and role="alert" on the error line. Declined: confirm-before-disconnect on the cloud-backup row. Disconnect was unconfirmed before this PR too, so adding a dialog is a behaviour change beyond the redesign rather than a fix to it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
5b0ca3d874 |
fix(copy): make K3 and year-end claims match what the code actually does (#1431)
* fix(copy): make K3, leasing and year-end claims match what the code does
Follow-up to the batch that removed the uppskjuten-skatt posting on
obeskattade reserver (K3 29.37 gross in juridisk person) and added the K2
asset-account gate. Six user-facing strings still described the old
behaviour or made claims the code cannot support.
1. Arsredovisning page: the K3 explainer promised an uppskjuten skatt-not
and a materiella anlaggningstillgangar-not in every K3 document. Both
are conditional (a 2240/8940 balance, assets in the register) and the
first is now absent in the normal case. The kassaflodesanalys is
dropped with a warning when it cannot be generated, so it is named
only when the document actually carries one.
2. Regelverk settings: kassaflodesanalys was presented as following from
K3. It follows from being ett storre foretag
(swedish-year-end-closing/references/reporting-and-filing.md:10,
legal-framework.md:42); the copy now says the product includes one and
states the storre-foretag rule separately. Komponentavskrivning was
presented as optional under K3; it is mandatory where component useful
lives differ materially (k2-vs-k3.md:5, asset-accounting
references/depreciation.md:33).
3. Note 1 and the Uppskjutna skatter-not no longer claim the 2240 balance
is hanforlig till obeskattade reserver. deriveLatentTaxMovement reads
the 2240/8940 balances only, and under K3 that account carries deferred
tax on all temporary differences (k2-vs-k3.md:11-13).
4. The deferredTax 'unknown' branch emitted the gross-reserve statement,
which is the denial phrased positively: the same affirmative claim
about books that could not be read. It now emits no deferred-tax
paragraph at all; build-data already warns on that path.
5. Capitalized-lease detection looked at 1260/1269 only. On the shipped
BAS 2026 chart 1260 is a free inventarier account and 1269 is ack.
avskrivningar pa datorer, so owned computers were reported as leased,
while 1217/1227 (finansiellt leasade) were missed. Detection now reads
the company's own account names in kontogrupp 12, which is where BAS
keeps capitalized leases (leasing-and-disposal.md:28) and which owned
inventarier on 1220 never matches. 1720 forutbetalda leasingavgifter
stays out: that is the operational treatment.
6a. gnubok_year_end_readiness listed FX revaluation as a blocker (it is a
warning) and omitted UNBOOKED_TRANSACTIONS, the common one. The
description now names every actionable blocker kind, within the
280-char budget, and a test pins it against YEAR_END_BLOCKER_KIND.
6b. companies.accounting_framework defaults to 'k2', so every enskild
firma hit the K2 asset gate and was handed a BFNAR 2016:10 punkt 10.4
citation plus a K3 remedy it cannot take: a sole trader prepares ett
forenklat arsbokslut, not an arsredovisning (legal-framework.md:29,
:48). entity_type now rides along on the companies read the routes
already do, and non-AB entities get wording with no citation and no
K3, keeping the 1090 remedy. The K1 counterpart of punkt 10.4 is not
sourced in the repo skills, so nothing was invented in its place.
* fix(copy): close the review findings on the copy-truth sweep
Three follow-ups from the source and code reviews. (1) The K2/K3 help text had upgraded a vague sentence into a definite boundary claim ('gransen gar vid <trosklar>'), which excludes the other routes into mandatory K3 that are live right now for this control's audience: noterade vardepapper, and from fiscal years starting after 2025-12-31 also utlandsk filial, kryptotillgangar, aktierelaterade ersattningar and fastighetsbolag. An AB in one of those categories would have read the sentence and stayed on a regelverk it may no longer use. (2) hasCapitalizedLeaseAsset compared per-side cumulative totals, so a lease acquired earlier and disposed this year still claimed the balance sheet carries a leased asset; it now compares the net balance. (3) The K3 warning enumerated a kassaflodesanalys the document may not contain, contradicting the newly conditional page copy on the same screen.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Jakob Wennberg
|
78a37f6396 |
fix(year-end): typed preflight blocker codes so remediation links render (#1420)
* fix(year-end): typed preflight blocker codes so remediation links render validateYearEndReadiness emits Swedish blocker strings but the wizard's BlockerRow matched English phrases, so no remediation link ever rendered, and the voucher-gap branch pointed at /bookkeeping/voucher-gaps which only exists as an API route. Blockers now carry stable machine codes end to end (YearEndBlockerCode on YearEndValidation.blockers, mirrored additively as blockerItems on BokslutReadinessReport); errors stays the plain string mirror so the v1 compliance check and MCP tool keep their exact shapes. BlockerRow matches on code and links only to pages that exist; the voucher-gap and dead-link branches are removed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(year-end): code the unbooked-transaction blockers #1414 added #1414 landed two new blockers in validateYearEndReadiness using the old errors.push style, which this branch had already renamed to a typed blockers array. Merging main left them referencing a variable that no longer exists. Converted both to the typed scheme: UNBOOKED_TRANSACTIONS (the safety guard that stops executeYearEndClosing from aborting at the step 7 lock AFTER the closing entry posted at step 4) and UNBOOKED_CHECK_FAILED (the fail-closed variant). Neither behaviour changes; both keep their Swedish wording verbatim. The MCP year_end_readiness classifier now routes on the stable YearEndBlockerCode instead of regexing the Swedish message, with the wording heuristic kept as a fallback for an unmapped or legacy English message. The public `kind` values are unchanged, so MCP consumers see the same output; both new codes map to 'unbooked_transactions' as before, since an agent reacts to "we could not tell" the same way it reacts to a real count. UNBOOKED_TRANSACTIONS gets a /transactions remediation link in the preflight step: that page is where a transaction is booked or marked private, the two remedies the message names. UNBOOKED_CHECK_FAILED gets none: the remedy is to re-run the check. --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
576ed4d290 |
fix(agent): let the assistant see answers the user gave in WhatsApp (#1425)
Two field findings from the first live receipts.
1. The assistant re-asked for information the user had already given.
The user answered the representation question in WhatsApp ("Elias
Karlsson från Canguro Media, Jakob Wennberg från Arcim"), the answer
was stored correctly on invoice_inbox_items.channel_context, and then
the in-app assistant said it could see no participant names and asked
for them again. The intent's inbox query selected only document_id and
extracted_data, and nothing under lib/agent/ read channel_context at
all. It is now selected, threaded onto each underlag as chat_answers,
and rendered into the prompt as "uppgivna av användaren" with an
explicit instruction that human answers outrank anything read off the
image and must never be re-asked. Also backfilled by document_id: a
receipt can reach the intent through the document paths without its
inbox row being matched to the transaction.
2. The representation question accepted half an answer in silence.
Naming participants but no purpose stored purpose=null and replied
"Tack!", leaving the deduction undocumented while looking complete.
Skatteverket wants both (BFL 5 kap 6-7 §). It now asks once, for the
missing half only, and keeps the question open so the reply routes
back to the same receipt. Anti-loop: the follow-up fires only when no
representation block exists yet, so a second incomplete answer is
taken as-is rather than nagging.
Tests cover the prompt half and the query half separately: the earlier
prompt tests injected chat_answers directly and would have stayed green
with the column still missing from the select, which is precisely how the
bug shipped. Both mutation-checked.
Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Jakob Wennberg
|
c84f951a5c |
fix(whatsapp-inbox): name both ways to route receipts for multi-company senders (#1424)
Field feedback after the first live receipt: the link confirmation told a multi-company sender only to set a default company in the panel, so the per-receipt path looked unsupported even though it is the one that actually runs when no default is set. Now it names both: set a default, or send one receipt at a time and answer the company question after each. The guidance also moves to its own paragraph. Run together with the AI disclosure it read as one sentence, which is how a real user came away believing the bot had called itself a "mänsklig AI-assistent". New copy test asserts the promise (both options present, disclosure kept in its own paragraph, single-company message unchanged) rather than the exact wording, so a rewrite stays free but a dropped option fails. Mutation-checked against the previous copy. Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
c0825e9bd2 |
fix(bokslut): surface unbooked transactions and AR/AP tie-outs in year-end preflight (#1414)
* fix(bokslut): surface unbooked transactions and AR/AP tie-outs in year-end preflight Two gaps in the year-end readiness layer: 1. Unbooked bank transactions were enforced only by lockPeriod, which runs at step 7 of executeYearEndClosing, AFTER the closing entry has posted at step 4. A period with unbooked transactions reported ready: true from gnubok_year_end_readiness and the wizard, then aborted mid-flow, leaving a posted closing entry on an unlocked, unclosed period. The readiness check now runs the same counter as the lock guard (countUnbookedInPeriod, so the number reconciles with the "att bokföra" badge) as a blocking error, failing closed if the check cannot run. The lockPeriod guard stays as defense in depth. The MCP classifier tags the new blocker as kind unbooked_transactions. 2. The Phase-1 avstamningar (kundreskontra vs 1510, leverantörsreskontra vs 2440) existed as reports (lib/reports/ar-reconciliation.ts, supplier-reconciliation.ts) but were wired only to the ledger report routes, never to the bokslut preflight. The readiness aggregator now runs both tie-outs and surfaces mismatches as warning-severity reminders with deep links, mirroring the bank-reconciliation reminder. Warnings only, never blockers: a difference can be legitimate (FX-settled partials). Skipped entirely for kontantmetod companies, where open invoices are deliberately not on 1510/2440 until the year-end conversion exists and the tie-out is permanently unreconciled by construction. Unconvertible-FX rows produce a "could not reconcile" message instead of a phantom difference. YearEndValidation gains an optional unbookedTransactionCount field; the v1 compliance endpoint and MCP readiness tool pick the new blocker up automatically since they share the same engine. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): classify the next-period-IB readiness blocker instead of kind other The blocker "Nästa räkenskapsperiod har redan ingående balanser bokförda" was the only validateYearEndReadiness error with no classifier regex, so it always surfaced as kind: 'other'. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(bokslut): log swallowed AR/AP tie-out failures in the readiness aggregator Compliance-review finding: a rejected tie-out produced no reminder and no log entry, making a failed avstämning control indistinguishable from a reconciled one. Still degrades to no reminder (advisory check), but the rejection reason is now traceable, mirroring the unbooked-transaction check's logging. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
88760ae6f6 |
fix(whatsapp-inbox): harden against adversarial review findings (#1342)
* fix(whatsapp-inbox): erase the WhatsApp channel on account deletion whatsapp_phone_links relied on the auth.users ON DELETE CASCADE, but Accounted never deletes auth.users: account deletion is anonymize_user_account plus a ~100-year ban that keeps the auth row as a tombstone, so the cascade never fires and nothing revokes the link. After erasure the link stayed active with a decryptable phone_enc, lookupActiveLink kept resolving the number, and every further inbound message was persisted with body_text and the verbatim raw_payload while the bot kept replying: GDPR Art 17 plus continued collection with no lawful basis. The RPC is re-created verbatim from 20260724150000 with one added block that revokes and crypto-shreds the link, resets its conversation, nulls body_text/raw_payload on that link's messages and deletes outstanding link codes, plus a guarded repair pass for tombstones anonymized before this migration. Covered by a pg-real test that fails against the previous definition. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(whatsapp-inbox): pepper the link-code hash and bound code minting hashLinkCode stored a bare sha256 over CODE_ALPHABET^6 = 30^6 values behind a fixed 'AC-' prefix. The module cited the invite-token pattern, but invite tokens are 256-bit random; this space enumerates offline in about a second, so hashing at rest protected nothing. The sibling phone-crypto.ts already states the team's own threat model for a LARGER space ("a plain sha256 would be brute-forceable ... hence the pepper"), so link codes now hash through the same env-mandated pepper. /link/start was also an authenticated unbounded INSERT that left every earlier code valid. Minting now burns the caller's unused codes (the code the panel shows is the only one that works) and is capped per TTL window, with the route answering 429 instead of throwing. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(whatsapp-inbox): harden the conversation layer against the review findings Pre-merge hardening of the unshipped chat layer. Every change below has a test that fails without it. Lifecycle and races: - conversation writes go through updateConversation(), an optimistic compare-and-set on updated_at (the trigger makes it a revision counter). The ack winner, the answer worker, the pin refresh and the sweep hold different claims, so blind whole-jsonb writes resurrected answered questions, wiped pending_question and dropped queue entries. - terminal markStatus writes are guarded on processing_status='processing' so a losing worker cannot overwrite the winner's 'done' and null its inbox_item_id. - the message -> inbox item path is idempotent: a pre-check plus a 23505 fallback adopt the item a concurrent worker created, instead of throwing after the WORM document is already committed. - PROCESSING_STUCK_MS 90s -> 5 min. The enforced step budget of one media row already exceeds 90s, so the sweep was re-claiming live workers. - sweep 2b re-arms only when the conversation itself has been quiet, not just the rows: pending_ack=false plus unacked rows is also the state of a live finalize, which produced a duplicate combined ack. - pin expiry re-checks against fresh state instead of writing back a stale whole context, which reverted company choices applied mid-pass. - askNextQueuedQuestion claims the pop before sending, so two answer workers cannot ask the same question twice. Company question: - the state is rolled back when the M6 send fails, so the next receipt re-asks instead of parking receipts behind a question nobody received. - applyCompanyChoice claims the open question (company_options) rather than the state: a double tap confirms once, a transient membership-query error is no longer read as "not a member", and a LATE answer still lands. - at the 48h TTL the parked receipts are kept, not discarded: options and staged rows survive so a late digit or tap still files them, and only rows past Meta's ~30-day media window get the terminal marker. - an out-of-range digit or a typed company name now gets the options repeated instead of silence or the "I cannot answer questions" reply. Inline dispositions: - stop/start/byt/company answers run their side effect BEFORE the terminal wamid row, with a SELECT pre-check for dedupe. Writing the row 'done' first made them at-most-once: a crash in between lost the action forever. Copy and answers: - acks state the extracted currency instead of labelling every total 'kr'. - M17 stops promising "about 10 minutes" when the daily quota tripped. - M18 is sent once per message tracked by the outbound row, so a file whose first attempt died still reaches the sender, including from the max-attempts path. - M11 no longer claims the number is disconnected: 'stopp' pauses, and muted senders now persist no chat content at all. - 'byt' is recognized in every state but awaiting_company (m6-confirm teaches the word, and it was being stored as answer data instead). - text sent while a re-send question is open is kept as a note on THAT receipt with the question left open, instead of binding to another receipt's question. - a quoted reply wins over the pending question and is appended when the quoted question is already answered, so corrections stop landing on the wrong receipt. - context answers keep raw_answer + answered_at like representation does. - finalizeBurst checks the send result: on failure it rolls the question back and leaves the rows unacked for the sweep. PII: - the sender's plaintext number is stripped from raw_payload before it is persisted; replies decrypt the link's phone_enc instead. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs(whatsapp-inbox): record the erasure path and the hardening decisions RoPA gains the account-deletion row (immediate, not via the cron: the auth.users cascade never fires because the row is tombstoned) plus the two new security measures, and its "never in the clear" phone claim is now true of the stored payload. DECISIONS.md records the non-obvious calls: revoke-not-delete on erasure, commit-then-roll-back for the company question, keeping expired company choices answerable, the compare-and-set conversation write, effect-before-terminal-row for inline dispositions, honest M11 copy, and the raw_payload redaction. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(whatsapp-inbox): stop the answer re-claim from following a confirm with M16 A worker that died after applying an answer and sending its confirmation leaves the row 'processing'. The sweep re-runs it, resolveAnswerTarget finds the question already answered, and the user got "I did not understand" immediately after the confirmation they had just received. The fallback is now first-attempt only. The catch comment claiming the sweep retries these rows is corrected too: 'error' is terminal for the sweep, and nothing on the answer path throws anyway (interpretChatAnswer degrades, sends never throw, supabase-js returns errors), so the catch is a programming-error net. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(whatsapp-inbox): drop the amount floor on the representation question The Swedish compliance review on #1340 caught a real error in the trigger rules: the representation question only fired above 150 kr, but the duty to document deltagare and syfte is what makes the expense deductible at all (BFL 5 kap 6-7 §) and it is not conditioned on any amount. The 300 kr per person figure I had in mind is the VAT-deduction base cap, a different rule. A 120 kr business lunch would have been booked with no participant trail, which is exactly the deduction Skatteverket denies later. Noise stays bounded by the triggers that were already there: the question fires only for receipt-shaped documents from restaurant, cafe or hotel merchants, at most once per receipt, twice per burst and six times per sender per day, and a single "nej" dismisses it. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> |
||
|
Jakob Wennberg
|
bf5ca2c615 |
feat(whatsapp-inbox): GDPR retention cron and RoPA entry (#1341)
PR5b, the final code piece of the WhatsApp intake track. A daily cron (04:15) enforces the channel's retention table; the receipt itself stays 7-year WORM under BFL and is never touched. Retention actions (lib/retention.ts, each isolated and idempotent): - whatsapp_messages transcripts past 90 days: body_text + raw_payload cleared in id batches under a wall-clock budget; the row skeleton (wamid, direction, timestamps, status, inbox_item_id) survives for audit. Only rows still carrying content match. - Rows with phone_link_id IS NULL (unknown senders, orphans) past 30 days: deleted. - Link codes expired more than 24h ago: deleted, used or not. - Sender rate counters idle 2+ days: deleted (minute/day window keys are dead weight after that). - Links revoked 90+ days ago: phone_enc crypto-shredded to '' (column is NOT NULL), one-shot via neq guard; phone_hash and phone_masked kept for uniqueness history and audit display. Route mirrors the sweep cron exactly: withCronContext + registry gate (503 EXTENSION_DISABLED when the extension is off). vercel.json gets the schedule and both Docker crontabs are regenerated. Compliance: new whatsapp.receipt_intake activity in .compliance/ropa.yaml covering purpose, Art 6(1)(b)/(c)/(f) bases with the Art 14(5)(b) note for third-party attendee names, Meta Platforms Ireland as processor (Cloud API, EU SCC addendum, Local Storage region DE), the differentiated retention table, and security measures. Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
629069e281 |
feat(whatsapp-inbox): conversation layer with clarifying questions (#1340)
PR4 of the WhatsApp intake track: turns the per-message PR3 pipeline into a
conversation. Media replies are burst-debounced into ONE combined ack (M4
single / M5 numbered list) sent by the single winner of the atomic
pending_ack claim; losers stay silent. Multi-company senders get the company
question (reply buttons <=3, list 4-10, numbered text >10) with an 8h
sliding pin ('byt' clears it); their receipts park as staged message rows
until the answer and then run through the normal intake path.
Clarifying questions are evaluated per receipt after extraction, max one per
receipt, priority unreadable > representation > partial, keyed on the
Phase-0 classification (legibility/documentKind/merchantCategory) with
heuristic fallbacks (compressed-chat-photo signal, extended meal regex).
Budgets: <=2 content questions per burst, <=6 per sender per Stockholm day;
over budget acks only and flags the item moved_to_app. Questions expire
after 48h (sweep, silent hand-off) and are asked exactly once.
Free-text answers route through the ONE new LLM call
(lib/interpret-answer.ts): Sonnet via Bedrock, max_tokens 600, no thinking,
forced tool call validated by Zod with hard caps, gated by
checkAgentRateLimit, reply framed as untrusted data. Any failure degrades to
storing the raw text as a note; exact 'nej' short-circuits without the LLM.
Answers land in invoice_inbox_items.channel_context
(representation/user_note/quality) with ChannelQuestionAsked/Answered
processing-history events. Late answers match by quoted wamid or the most
recent open question within 7 days.
New per-minute sweep cron (registry-gated physical route, 503
EXTENSION_DISABLED when off) re-claims stuck rows (max 3 attempts), rescues
crashed burst acks, expires questions and pins. One new migration
(20260802210000) adds whatsapp_messages.acked_at, the relational burst-
membership marker, with pg-real coverage for the single-winner claim.
Verified: full vitest suite (12270), pg-real against a migrated
supabase/postgres 15 (977), lint 0 errors, tsc at the 405 baseline,
check:guards green, crontabs regenerated. Mutation-checked the debounce
claim and the daily budget gate.
Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Jakob Wennberg
|
d1c411ad6f |
feat(invoice-inbox): surface WhatsApp chat context in booking flows (#1339)
* feat(invoice-inbox): surface WhatsApp chat context in booking flows
The WhatsApp intake bot writes verified human answers (photo caption,
representation deltagare + syfte, sender note, open-question state) to
invoice_inbox_items.channel_context. This makes the in-app booking flows
READ it:
- New core renderer lib/documents/channel-context-notes.ts: deterministic
compact Swedish line ("Representation: Anna Berg (Volvo), Jakob W ·
Syfte: uppföljning av avtal"), capped at 220 chars by dropping whole
participant names ("… och N till"), never mid-name. Representation
first, then user_note; caption only when nothing else exists.
- FieldsRail "Från WhatsApp" block in InvoiceInboxWorkspace: caption,
deltagare, syfte, anteckning rows plus an ochre AttnLine when a chat
question expired unanswered (pending_question.status = moved_to_app).
- Notes threading: book-direct and convert default their notes to the
rendered string server-side when the request carries none (a supplied
value always wins); BookDirectlyDialog prefills its notes input with
the same string so the user can edit it before it lands. Bulk-book
(categorize-core) joins the shared batch note with the per-item
rendered context so the representation trail survives batch booking.
- Inbox list: whatsapp rows get a chat icon and a quiet "Fråga obesvarad"
badge for moved_to_app items. No worklist count change: unresolved
whatsapp items are already counted by countInboxDocuments.
- sv/en strings for every new key; renderer + route + bulk tests.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(invoice-inbox): honor cleared notes and keep unreviewed captions out of verifikat
Two adversarial-review findings on the WhatsApp surfacing flows, both about
text that lands on an immutable verifikat.
1. A cleared note was silently re-applied. BookDirectlyDialog prefills the
rendered chat context, and the dialog sent `notes.trim() || undefined`
while book-direct and convert defaulted from channel_context on any falsy
value. A user who read the prefill, disagreed and deleted it therefore got
it written back onto a posted entry, removable only through a formal
rättelse. Both code comments claimed "an edited value always wins", which
was false for exactly that edit. Now PRESENCE of the field decides: the
dialog always submits `notes` (empty string included) and the routes only
default when the field is absent from the request (MCP, older clients).
The Zod `.optional()` carrying that distinction is documented at the
schema so it is not "tidied" into a `.default('')` later.
2. The photo caption was auto-burned into verifikat text with no review.
renderChannelContextNotes fell back to the raw caption, and bulk-book
appended the result per item with no per-item notes field at all (the MCP
approval preview deliberately shows no per-item PII either), so unreviewed
chat text reached a WORM record nobody had seen. The renderer now takes
{ includeCaption } and leaves the caption out by DEFAULT: representation
answers and user_note are replies to a question the bot asked, the caption
is not. Only the Bokför direkt prefill opts in, where the user reads the
string in an editable field before booking.
Tests: cleared-notes and whitespace-cleared on book-direct, cleared-notes on
convert, caption-never-defaulted on both routes, caption-not-threaded in
bulk-book, and the renderer's opt-in. The book-direct cleared-notes tests
were mutation-checked (restoring the truthiness fallback fails them).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(reports): export the chat answers behind a verifikat in the full archive
The verifikat line caps the representation trail at 220 chars and drops whole
participant names ("… och N till"); the complete list (deltagare, syfte,
raw_answer) exists only in invoice_inbox_items.channel_context. That table was
in ARCHIVE_EXCLUDED_TABLES with a rationale predating channel_context ("inbox
workflow state"), so a company leaving Accounted and keeping the full-archive
export as its BFL 7-year record kept an incomplete deltagare documentation for
its representation deductions.
Dumped as a column PROJECTION, not the whole row: the new
MasterDataTableSpec.columns narrows the select to the underlag provenance
(document, matched transaction, created verifikat / leverantörsfaktura) plus
channel_context, so the answers are tied to what was booked from them while
the inbox workflow state (email bodies, OCR output, error messages) stays out
of the archive. The documents themselves remain in dokument/.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Jakob Wennberg
|
398c734b93 |
feat(whatsapp-inbox): intake extension with webhook, phone linking and receipt ack (#1338)
Webhook lifecycle: GET hub.challenge handshake (constant-time verify-token compare); POST verifies X-Hub-Signature-256 over the RAW body before any parse, Zod-parses the envelope, persists inbound rows (partial-unique wamid = dedupe against Meta's up-to-7-day redelivery), acks 200 fast and defers media processing via the after() idiom. Rejected and rate-limited content always acks 200 and lands as skipped/error rows, never a retryable status. Linking: the settings panel (Installningar -> WhatsApp) mints AC- one-time codes (sha256 stored, 10 min TTL, single use, ambiguity-free alphabet); the webhook consumes the code, binds phone to user (HMAC-peppered hash + AES-256- GCM at rest) and confirms with M3. Keyword commands stopp/start/hjalp; unknown senders get one throttled M1 greeting (1/h, 3/day) behind the sender-quota RPC, with no media download and no content persistence. Intake worker: atomic claim on the message row (the durable job record), company resolution (default -> sole membership -> M6 fallback, no item), per-company inbox quota (ack-and-drop, M17 once per 10 min per sender), MIME allowlist, 10 MB stream-checked media download, exact sha256 duplicate check, then the shared uploadAndExtract funnel (source 'whatsapp', channel_context caption, whatsapp_message_id) and the M4 ack with extracted merchant/total/date. Failures wrap to 'error' + error_message + one M18. uploadAndExtract widened: source 'whatsapp', optional channelMeta + actorId; email/upload paths behaviorally unchanged. Deferred to PR4: burst debounce + combined ack (M5), in-chat company choice (M6 buttons + 8h pin), clarifying questions M7-M10, interpret-answer LLM call, sweep cron, retention cron. Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
02e48efe11 |
refactor(invoice-inbox): export uploadAndExtract from lib (#1336)
Pure move of uploadAndExtract and its private helpers (sanitisers, page-count/slice, sandbox check, MIME/size consts) from index.ts into lib/upload-and-extract.ts so a future channel extension can import the shared funnel the way document-extraction already imports extract-invoice-fields. No behavior change; existing tests prove it. Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
3b3adf96c7 |
feat(invoice-inbox): receipt-aware extraction + OCR pipeline fixes (#1331)
Receipts and invoices were extracted through one invoice-shaped prompt with no document classification. The extractor now also returns documentKind, payment method (+ card last4), purchaseTime, merchantCategory and legibility, validated with .catch(null) so a hallucinated label degrades to unknown instead of sinking the parse. The FieldsRail shows type and payment method above the editable fields. Pipeline fixes, all verified against real failure paths: - PDFs >3 pages: extract from a pdf-lib slice of the first 3 pages instead of skipping entirely (issue #553 gate); truncation recorded in extracted_data.pages and shown in the UI. - Oversized images (>4 MB, over Bedrock's 5 MB cap): downscale to <=2000px JPEG via sharp before base64, instead of erroring to an empty result. - HEIC/HEIF: attempt sharp transcode to JPEG; when libvips lacks HEIF (prebuilt binaries), fall through to today's behavior but show an explicit hint instead of silently blank fields. - Oresavrundning: prompt rule + totals.roundingAmount so receipt totals reconcile with subtotal+VAT for exact-amount transaction matching. - retry-extraction overwrites extracted_data wholesale: a confirm now guards against silently destroying manual field edits. Deliberately NOT added: retry-on-transient-Bedrock-error; the SDK already retries twice by default (maxRetries=2). Co-authored-by: Jakob Wennberg <jakob.wennberg@gmail.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
9f5a43310b |
fix(salary): recompute entitled_days on existing ledger rows and record pre-cutover taken days (#1403)
The vacation ledger sync carried entitled_days verbatim on existing open rows while re-deriving accrued and taken, so a stale entitled value (for example the flat 25 stored before Semesterlagen 7 § pro-rating existed) survived every sync. The recompute loop now re-derives entitled the same way the lazy-seed path does, with the opening-balance cutover still outranking recomputation for the year containing cutover_date. Opening balances could also not record paid vacation days already taken in the cutover year under the previous payroll system. New additive column employee_opening_balances.vacation_days_taken_this_year (NUMERIC NOT NULL DEFAULT 0, CHECK 0..40) threaded through the shared service, the Zod schema, the MCP staging tool (schema + mergeable fields), the staged-operation executor, the v1 REST routes, and the employee editor form. Ledger semantics for the cutover year, on both seed and recompute paths: entitled = remaining + taken_this_year, taken = booked-run taken + taken_this_year, so remaining keeps meaning remaining and the seeded value survives every subsequent sync. Fixes #1347 Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
00ae3540db |
feat(customers): carry contact person and invoice copy recipients through migration (#1392)
* feat(customers): carry contact person and invoice copy recipients through migration Extends the arcim-migration entity mapper, Fortnox provider mapper, canonical DTOs, customer APIs (web + v1) and invoice send flows so contact person and customer-level invoice CC/BCC addresses survive provider migrations. NULL means unconfigured and empty means an explicit clear, so re-syncs enrich legacy gaps without resurrecting deliberately removed values. Fortnox fixed assets are split into a dedicated follow-up issue. Fixes #1345 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * chore(db): bump customer metadata migration past pack-slug version Main already contains 20260803230000; keep new versions strictly newest so Supabase branching applies them in order. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(customers): complete Customer type consumers and make enrichment payload resolvable The preview-pdf mock customer and the makeCustomer fixture now carry the three new metadata fields, fixing the type-check failure in Build (zero extensions) and Vercel. The enrichment update in the migration orchestrator now spells its payload as an object literal typed CustomerMetadataEnrichment (absent keys drop at serialization), so the phantom-column guard resolves the columns instead of counting another unresolvable dynamic payload past its ceiling. The cc/bcc guards also verify element types instead of casting. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
267ed6c1bb |
feat(mcp): expose the konteringspaket catalogue as a resource (#1395)
An agent proposing a booking currently invents the account numbers, and a plausible guess (6071 instead of 6072, or the full cost instead of the 80% deductible share) produces a verifikat that posts and is wrong. This turns that into naming a reviewed template. Read from packs/ rather than the database on purpose: legal_note has no column in booking_template_library, so the database copy cannot answer "when does this template apply", and that note is the one field an agent cannot derive from account numbers. Carries the amount maths explicitly (vat lines from vat_rate, everything else from ratio) with an instruction not to fudge an amount to force a balance, plus notes that entity_type is binding, that a company may hold templates beyond this list, and that posting goes through the journal-entry tools so period locks and the balance check still apply. A load failure returns an explicit error rather than an empty list: an agent shown 12 of 26 templates concludes the other 14 do not exist and hand-rolls accounts for them. Verified the packs are actually reachable at runtime before relying on this: a production build traces all 26 YAML files into the serverless function, so the loader is not a local-only convenience. Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Mattsson
|
281cb3989d |
fix(build): break Turbopack chunk-name hash collision from #1385's import edge (#1394)
Since the #1385 squash-merge every production build failed with 'Two or
more assets with different content were emitted to the same output path'
on [root-of-the-server]__1ge0sz5._.js: two distinct server chunk groups
(an AWS smithy helper chunk and the withRouteContext auth chunk) hash to
the same chunk name. The graph change that tipped the chunk layout into
the colliding state was entity-mapper.ts (arcim-migration extension
entry graph) importing lib/vat/supplier-invoice-line-checks, which
drags lib/money into the extension root.
Move normalizeVatRateToFraction into an import-free leaf module
(lib/vat/vat-rate-unit.ts), re-export it from
supplier-invoice-line-checks for all existing callers, and point
entity-mapper at the leaf. Behavior is unchanged (511 targeted tests
pass); the server chunk graph returns to the pre-#1385 shape that
builds cleanly. Verified: npm run build fails on
|
||
|
Mattsson
|
8443062b1f |
fix(vat): enforce fraction unit for supplier invoice vat_rate writes (#1385)
Closes the remaining #310 write paths: credit-note item copies (web, v1, pending-operations) and arcim-migration supplier imports now normalize vat_rate to the decimal-fraction unit before storage, and a NOT VALID CHECK constraint guards every new supplier_invoice_items row. Customer invoice items deliberately stay percent; legacy supplier rows are left untouched so posted-entry reversals reuse the exact original values. Fixes #310 Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
8299ee9fb4 |
fix(bookkeeping): resolve settlement account in all categorization flows and ship mis-booking audit (#1383)
Completes the #985/#986/#987 caller sweep: categorize-core, v1 batch-categorize, pending-operation edits and the MCP categorize path now resolve the settlement leg from the transaction's cash account instead of inheriting a hardcoded or stale account. Extends the correct_entry preview with currency, tax and dimension line metadata so staged corrections preserve full line fidelity. Adds a read-only audit query and a runbook for reviewing and correcting historical mis-bookings via staged storno with explicit approval; no automated bulk mutation. Fixes #1001 Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
5d7952a01e |
feat(mcp): model-free document upload via signed URL (#1378)
* feat(mcp): model-free document upload via signed URL (#748) Adds gnubok_create_document_upload + gnubok_complete_document_upload so document bytes reach storage through a short-lived signed PUT URL and never pass through the model context. Fixes silent base64 corruption on real-size PDFs and the context blowup on batch uploads. - pending/ staage keys with TTL cleanup; completion validates magic bytes + SHA-256, moves bytes to the WORM key and adopts the reserved UUID as document id, making retries and concurrent completions idempotent - legacy gnubok_upload_document kept for clients without file access, description now points to the signed-URL pair; shared mime resolution and inbox-item creation extracted - both new tools mapped in TOOL_SCOPE_MAP (transactions:write) and MCP_TOOL_CAPABILITY_MAP (ai) so the paywall and scope gates hold - payload guard ceiling 58.5K to 59K after trimming the create tool's outputSchema to upload_id/upload_url/expires_at Fixes #748 Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): satisfy capability-map lock and phantom-column scanner The exact-entries lock in capability-maps.test.ts now includes the signed-URL pair as dispatch-only AI tools, and the inbox insert uses a literal payload (explicit UUID instead of a conditional spread) so the no-phantom-columns scanner can resolve every column. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
cd7d7f52b9 |
feat(invoices): per-recipient email delivery outcomes (#1384)
* feat(invoices): per-recipient email delivery outcomes Resend delivery webhooks identify affected addresses in data.to, so one message with CC recipients can carry independent To/CC outcomes instead of masking the failing address into the aggregate reason text. - new apply_invoice_delivery_provider_event RPC merges each reported recipient onto its immutable To/CC position with the same rank and timestamp ordering as the aggregate status (retry and out-of-order safe) - recipient map is PII-free: keyed to:N / cc:N, BCC and unmatched recipients are never represented, and the map is cleared on PII redaction - delivery summaries, API route and MCP tool expose the sanitized map; the route re-sanitizes as defense in depth - UI shows a per-recipient status list under the aggregate outcome The prod ops check in issue #1350 (webhook registered in Resend and RESEND_DELIVERY_WEBHOOK_SECRET set in Vercel) cannot be verified from the repo and remains a follow-up. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * test(invoices): commit provider event before cross-context read The BCC-leak test applied the event inside the rollback-scoped service role helper and then asserted through a separate member context, so the applied status was rolled back before the read. Use the committing runAsServiceRole helper for the apply, matching how the summary read is performed in its own context. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Mattsson
|
8a2498a987 |
fix(mcp): add true continuation to invoice list tools (#1327)
* fix(mcp): add true list continuation Signed-off-by: Emil <emilmattsson14@gmail.com> * test(mcp): cover pagination edge cases Signed-off-by: Emil <emilmattsson14@gmail.com> * fix(mcp): validate pagination offsets Signed-off-by: Emil <emilmattsson14@gmail.com> * fix(mcp): preserve continuation without counts Signed-off-by: Emil <emilmattsson14@gmail.com> --------- Signed-off-by: Emil <emilmattsson14@gmail.com> |
||
|
Mattsson
|
bfbd926950 |
fix: paginate MCP inbox items (#1329)
* fix: paginate MCP inbox items * fix: validate MCP inbox cursors * test: reset MCP inbox pagination state |
||
|
Mattsson
|
bcf919d995 |
fix: expose inbox filenames via MCP (#1328)
* fix: expose inbox filenames via MCP * fix: handle joined inbox attachments |
||
|
Mattsson
|
fd376eff94 |
fix: audit Cloud Backup OAuth redirects (#1324)
* fix(cloud-backup): pin OAuth callback origin * fix: reject non-web cloud backup origins |
||
|
Jakob Wennberg
|
0f1c7c9365 |
fix(import): refuse a Bokio connection that opens a different company (#1315)
A Bokio integration token is scoped to one Bokio company and the company id is typed in by hand, so credentials for the user's other company imported that company's customers, suppliers and invoices with no error at all. Probe /companies/{id} before storing, mirroring the Bjorn Lunden /details probe, and refuse on a confident org-number mismatch.
Also surface the inbox mail body when nothing was attached: it was captured in email_body_text and never read back, which made Gmail's forwarding-confirmation mail unreadable and the forward impossible to complete.
|
||
|
Jakob Wennberg
|
4a38fa30ed |
fix(recon): share the cash-account scope and cover the no-1930 case (#1309)
PR #1295 (
|
||
|
Mattsson
|
144cc51458 |
fix(mcp): scope VAT close reconciliation account (#1295)
Resolve the VAT close reconciliation scope using the selected cash account, currency, and unassigned-transaction behavior. Add regression coverage for cross-account leakage and fail closed on lookup errors.\n\nCloses #1290 |
||
|
Mattsson
|
6318501b71 | fix(vat): recover ruta 05 for null-rate custom accounts (#1296) | ||
|
Mattsson
|
17a7a62ceb |
fix(reports): stop the resultatavslut zeroing declarations, and make the mistake uninventable (#1293)
* fix(settings): explain why account deletion is blocked The delete-account button was disabled while the user still owned companies, but the reason only lived behind the "?" on the blocker row, so the greyed-out button read as broken. Surface it as one visible attn sentence directly under the button, and point aria-describedby at it whenever the button is disabled, not only on a load error. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(enable-banking): share one PSD2 consent across a user's companies Connecting the same bank for a second company required a second BankID, and at SEB that new authorization silently revoked the first one. A user with four companies at one bank therefore signed four times a quarter and ended up with three dead feeds, each still rendering as "Aktiv" with a stale last_synced_at until someone pressed Synka. Prod says this is not one customer: every SEB customer holding connections in more than one company has had an earlier company stop syncing at the moment the next was authorized, most of them while the consent was still formally valid for weeks. The same measurement over other banks is far quieter, so the one-active-session-per-PSU limit is real and ASPSP-side. Enable Banking already supports the shape we want. POST /auth carries no account restriction, so a session covers every account the user ticked at the bank, and GET /accounts/{uid}/transactions takes no session id, so a second company can sync its own accounts from an existing session. bank_connections has no unique constraint on session_id, so this needs no migration. Adds lib/session-sharing.ts plus GET /reusable-sessions and POST /attach. When a live session in another of the user's companies still exposes accounts no company syncs, the settings panel offers to reuse it: the new row shares session_id and consent_expires, carries only the unclaimed accounts, and lands in pending_selection so the existing IBAN-aware account picker does the ledger mapping. Only the consent is shared; accounts, cash_accounts and transactions stay strictly per-company. Sharing a session changes three lifecycle paths, all handled here: - Disconnect and reconnect now refcount before revoking. A blind revoke would take down a sibling company's feed, which is the exact failure this removes. The count runs on a service-role client because RLS hides a sibling in a company the user has since left, and it fails closed: an uncertain count is treated as shared, since a lingering consent lapses on its own in 90 days while a wrongly revoked one kills a working feed. - A renewed consent fans out to every company sharing the old session, and re-points their account uids by IBAN. Several ASPSPs reissue uids on re-authorization, so carrying the session id alone would have left siblings calling retired uids and re-broken them every quarter. This is also why the superseded session_id is no longer nulled at /connect: the callback needs it. - The nightly probe runs once per distinct session and applies the verdict to every row holding it, and expiry mails are keyed per (user, session), so one dead consent is one probe and one mail rather than four of each. Only enabled cash_accounts rows count as claiming an IBAN. The callback mirrors every account in a consent, deselected ones included, so counting any row as a claim would leave nothing offerable once the first company connects. An account handed to a company also stops being offered while that company's picker is still open, closing the window where two companies could book the same physical account. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(ink2): read the resultaträkning from the pre-closing books INK2R summed journal entries raw, so it included the resultatavslut that zeroes every P&L account into 2099 at year-end. Nettoomsättning, kostnader, periodiseringsfond and skatt all came out as 0, which cascaded into INK2S 7650/7651 and the taxable result. INK2 is always filed after bokslut, so this was every real declaration, and nothing warned: with the P&L at zero the balance sheet still tied out. INK2R now reads two views of the same period. The balance sheet comes from the closed books so 7302 keeps arets resultat via 2099; the income statement comes from the pre-closing books via excludeFinalClosingEntry, which drops only fiscal_periods.closing_entry_id so skatt and bokslutsdispositioner stay on the form (7525, 7528). The equity adjustment is now conditional on a posted closing entry having moved the result into 2099. Second, independent bug: accounts were mapped by BAS number with no regard for the sign of the balance, so konto 1630 with a credit was reported as a negative fordran instead of a skatteskuld and konto 2641 with a debit was netted off the liabilities. The three sign-reclassification rules the K2 iXBRL mapper already had are extracted to lib/reports/sign-reclassification .ts and applied to INK2R too, so both statutory reports present the same balance sheet. Only the rule table is shared: k2-mapper keeps its sumOre arithmetic because the iXBRL path is ore-exact while INK2R truncates per SFL 22:1. NE-bilaga had the same empty-resultatrakning bug and gets the same fix. Adds the closed-period coverage that was missing: the old tests only exercised the mapping table against an open period, the one state in which the engine happened to work. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(reports): make the year-end closing decision explicit at every call site generateTrialBalance took two optional booleans, so a caller that never thought about the resultatavslut silently got 'include'. That is the wrong default for anything summing class 3-8: the closing verifikat posts the mirror image of every P&L account into 2099 inside the same period, so the report reads ZERO across the board while the balance sheet still ties out and nothing warns. The booleans are replaced by a required closingEntry: 'include' | 'exclude-final' | 'exclude-all-year-end' with no default, so the build fails until each call site decides. All 40 were audited individually; every one keeps its current behaviour except the two that were provably broken: - Resultatrapport read zero on every line for a closed year, in JSON, PDF and XLSX, and its prior-year comparison column read zero for anyone whose previous year was closed. - Resultat per projekt (dimension-pnl) had the same defect and must stay in lockstep with Resultatrapport to keep reconciling. Both now pass 'exclude-all-year-end', which keeps them agreeing with the formal Resultaträkning rather than pre-empting Stage 2 of #1051 (DECISIONS.md:632). Deliberately unchanged and recorded in DECISIONS.md: the KPI expense composition, which is blank for a closed year but cannot be fixed without a migration and a displayed-figure change, and getBookedBolagsskatt, whose contract is an open period and whose call chain already caused a too-high-tax customer bug once. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(vat): keep the resultatavslut out of the momsdeklaration The closing verifikat posts the mirror image of every P&L account into 2099 inside the same fiscal period. Revenue accounts drive rutor 05, 39 and 40, so any VAT period containing the fiscal-year end reported NEGATED turnover once the year was closed. get_vat_declaration_totals already excluded vat_settlement and opening_balance entries, but not this one. Reproduced read-only against production: for December of a closed year the December declaration reported ruta 39 = -794 734 kr. After the fix that period reports 0 and the January period carrying the real sale is unchanged at 794 734 kr. Keyed on fiscal_periods.closing_entry_id, not source_type = 'year_end': avskrivningar, periodiseringsfond and skatt share that source_type and must keep whatever VAT effect they carry. A reversed closing entry is retained together with its storno so the pair still nets to zero, the same predicate trial-balance.ts uses for closingEntry: 'exclude-final'. Migration applied to the staging branch only; prod gets it via merge. The pg test is written but has NOT been executed locally (no DATABASE_URL configured and no local Postgres), so CI is its first real run. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(kpi): keep the resultatavslut off the monthly chart The monthly income/expense chart summed every posted entry in the fiscal period. The closing verifikat posts the mirror image of every P&L account, so once a year was closed the fiscal-year-end month charted the whole year's revenue as negative income. Measured read-only on production: 28 companies across 34 month-rows. The worst case charted December income as -10 347 459,81 kr where the real figure is +12,88 kr. Other examples: -1 868 731 -> +128 730, -1 850 501 -> +431 709. Both paths are fixed together so they keep agreeing: the RPC's monthly section now joins the tb_ex_ye_entries CTE it already computes for tb_ex_year_end, and monthly-breakdown.ts (the dimension-filtered fallback and the MCP path) gains the matching source_type filter plus the storno/correction chain of REVERSED year-end entries, so an undone bokslut does not leave half a pair behind. Migration 20260723180000 had recorded the omission as deliberate, on the grounds that it mirrored the JS scan. It did, but the JS scan was wrong. Migration applied to the staging branch (function body identical; three comment lines differ from the committed file). Prod gets the file via merge. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test(reports): pin every statement generator against a closed fiscal year The per-generator suites all exercised an OPEN fiscal period, which is the one state in which a generator that forgets the resultatavslut happens to work. Declarations are filed AFTER bokslut, so the untested state was the only state that occurs in production. That is why the same defect could ship three times. Two new suites over one shared fixture (closed-year-fixture.ts, a synthetic closed AB with a resultatavslut, a credit 1630 and a debit 2641): closed-year-statements.test.ts enumerates the generators and asserts each reports the year's revenue rather than zero, plus its own bottom line. The table IS the checklist: a new report either appears in it or nothing stops it shipping with this bug. Verified by regressing income-statement back to closingEntry 'include', which fails 2 of its assertions. cross-surface-agreement.test.ts asserts the surfaces agree with each other, which is what every customer complaint actually was. INK2R and the K2 årsredovisning must produce the same årets resultat, the same fritt eget kapital, the same sign reclassifications and the same balance total. The operational family (Resultaträkning, Resultatrapport) must agree internally, and the gap BETWEEN the families is asserted explicitly as bokslutsdispositioner + skatt, so when Stage 2 of #1051 lands the test names the expectation to change instead of failing vaguely. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * chore(guards): ratchet against new reports that scan the ledger directly A statement generator that aggregates journal_entry_lines itself has to remember, on its own, that the resultatavslut posts the mirror image of every P&L account into 2099 inside the same fiscal period. Three forgot, and each read ZERO revenue for a closed year while the balance sheet still tied out, so nothing warned. generateTrialBalance now requires an explicit closingEntry mode, which makes that decision a compile error. This guard is what keeps NEW reports on that path: any generator under lib/reports or lib/bokslut that reads journal_entry_lines and is not in the baseline set fails CI. Verified by adding a throwaway report, which the guard rejects by name. Voucher and line listings (general-ledger, journal-register, SIE export, reconciliation, diagnostics) are sanctioned: they show the ledger as posted and have no closingEntry decision to make. Four existing lib/bokslut files are grandfathered rather than migrated. One of them is a genuine open follow-up recorded in DECISIONS.md: sarskild-loneskatt-calculator sums 7410-7419 with no year-end exclusion, so its basis reads ~0 if it runs against an already-closed period. Left alone deliberately: it is a tax figure whose call chain has caused a customer bug before and deserves its own verified change. Also ratchets naive-ore-round down 646 -> 641. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * test(reports): pin where sign reclassification applies, in both directions No behaviour change. The sweep asked whether the 1630/2641 sign reclassification should be extended to the remaining balance-sheet surfaces; the answer is that there are none left. Both STATUTORY presentations already have it: the K2 iXBRL årsredovisning since 2026-07-23 and INK2R since 2026-07-29. The other two balance-sheet surfaces must NOT have it: /rapporter Balansräkning and Balansrapport are organised by account number under BAS-prefix headings, and balansrapport documents an invariant that depends on every row staying debit-positive where it was booked. Moving konto 1630 into a liability section would break the add-the-rows-to-verify-the-balance property and hide the account from anyone looking it up by number. Asserting both halves is the point. The first half stops the reclassification silently disappearing from one statutory surface again, which is how a customer ended up comparing two of our own reports against each other. The second half stops a future sweep "fixing" the operational reports into disagreeing with their own documented contract. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * feat(reports): detect statement disagreement instead of waiting for a customer Every year-end problem reported so far was a DISAGREEMENT between two of our own screens, not a single wrong screen. The årsredovisning said one figure, INK2 said another, and the customer did the reconciliation for us. Nothing in the product noticed, because each screen tied out on its own. Two additions: INK2R self-checks. On a closed year it compares the årets resultat it is about to declare against the booked konto 2099, and warns in Swedish when they disagree. This is the alarm that was missing: when INK2R reported 0 kr against a booked 469 542 kr, the balance sheet still balanced, so no warning fired. Mirrors the equivalent check k2-mapper has had since 2026-07-23, so both statutory reports now catch the same fault. reconcileStatements + GET /api/reports/statement-reconciliation return årets resultat from every surface side by side, grouped into families. ledger + statutory must agree and a mismatch is named; operational legitimately differs by bokslutsdispositioner + skatt until Stage 2 of #1051 lands, so that gap is explained rather than flagged. The visual panel is deliberately not built here: it needs a /frontend-design pass against the locked concept conventions plus sv/en strings, and the warning above already puts the alarm where the user looks. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(reports): address review findings from PR #1293 pg-real (7 failures, one signature): the new fixture called insertFiscalPeriod({ isClosed: true }) and then inserted journal entries into it, so enforce_period_lock (migration 017, legally required) refused the write. Not worked around: the RPC's predicate keys on fiscal_periods.closing_entry_id and never reads is_closed, so the fixture now links the closing entry and leaves the period open, which exercises the path that actually matters. CodeRabbit, closed-year-fixture: EX_YEAR_END_ROWS dropped only the P&L legs of the year_end entries (8811, 8910) and left their balance-sheet legs (2125, 2512) at pre-closing values, so the 'exclude-all-year-end' view sat 160 000 kr out of balance and misrepresented what generateTrialBalance returns. Latent, because today's consumers read class 3-8 only, but a shared fixture that does not balance is a trap for the next consumer. Both legs now go, and a new test asserts all three views sum to zero. CodeRabbit, INK2 totals: renamed totals.resultAfterFinancial to aretsResultat. It holds the result after bokslutsdispositioner AND skatt, which is årets resultat, not resultat efter finansiella poster, and build-data.ts uses the old name correctly for the different subtotal. The UI already labelled the value "Årets resultat", so the name was simply wrong. CodeRabbit, statement-reconciliation: the statutory branch called a generator and caught any throw as "wrong entity type", mapping genuine failures to a null figure that the comparison then skipped, so a real bug in a declaration generator made the function report isReconciled: true. That is the opposite of its purpose. It now dispatches on entity_type and surfaces a generation failure as a named disagreement. CodeRabbit, enable-banking (Emil's call to include): fetchClaimedIbans returned an empty Set on a cash_accounts read failure, which is indistinguishable from "nothing is claimed" and made every IBAN in the session offerable, including accounts another company already books to. Its own comment said it failed closed and its log said "offering nothing"; it failed open. Returns null now, and findReusableSessions offers nothing when the claimed set is unavailable. The test that pinned the fail-open asserted toHaveLength(1) under the name "offers nothing"; it now asserts []. Also removed an em dash per CLAUDE.md. The remaining enable-banking finding (consent-expiry cooldown stamped only on the selected connection, so it leaks one duplicate mail per sibling company) is deliberately left to Emil: it changes email-sending behaviour in his feature rather than fixing a stated contract. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> * fix(reports): resolve second-round review findings on PR #1293 pg-real, two NEW signatures (the closed-period one from cycle 1 is gone): kpi-report-aggregates-rpc.pg.test.ts asserted the exact contract migration 20260730090000 deliberately changes. Its comment read "year_end entries are NOT excluded from monthly" and expected December expenses 1250. That fixture's December holds only year-end-chain entries, so with the fix the month drops out of the chart entirely, which is the correct operational view: a month whose only activity is bokslut has no operating result. Assertion and file docstring updated to the new contract rather than the test being removed. vat-totals-closing-entry.pg.test.ts passed the wrong account arrays. p_net_ accounts is VAT_SETTLEMENT_NET_ACCOUNTS (2650/1650, the momsredovisning settlement pair), not the output-VAT accounts. Putting 2611 there made the extra year_end entry match the settlement-SHAPE detector, so an ordinary sale-with-VAT was classified a momsredovisning and dropped, and the test read 0 instead of 10 000. The RPC was right; the fixture was not. CodeRabbit, statement-reconciliation: resolveEntityType checked neither query's error, so a genuine DB failure (RLS, permissions, connectivity) returned null indistinguishably from "no entity type set", fell into the unsupported-form branch and reported isReconciled: true. That is the same silent-false-reconciled bug the cycle-1 refactor closed, one level down. The companies error now throws; a missing company_settings ROW stays tolerated, because .single() errors on zero rows and many companies have none. Mirrors the pattern the INK2 and NE engines already use. Still open by Emil's explicit choice: the consent-expiry cooldown is stamped only on the connection it was handed, so it leaks one duplicate mail per sibling company on the shared session. That changes email-sending behaviour in his feature rather than fixing a stated contract, so it stays his. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
16f34fb214 |
fix(arcim): resolve OAuth redirect_uri identically in authorize and exchange (#1287)
* fix(arcim): resolve OAuth redirect_uri identically in authorize and exchange The authorize leg honored the FORTNOX_REDIRECT_URI / VISMA_REDIRECT_URI override while the token-exchange leg hardcoded the NEXT_PUBLIC_APP_URL fallback. After the app-domain cutover (2026-07-21) the Fortnox env var still pointed at app.gnubok.se while NEXT_PUBLIC_APP_URL moved to app.accounted.se, so the two redirect_uri values differed and Fortnox rejected every code exchange (RFC 6749 4.1.3). The failure was invisible: the error popup posted its message from the old-domain origin, the wizard's event.origin check dropped it, and the popup closed itself. Both legs now resolve through one resolveArcimCallbackUrl() helper, and the error popup stays open with the reason on screen so a dropped postMessage can never again turn into "nothing happens". Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: log OAuth popup and redirect-uri rollout decisions Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
ef25a87d75 |
feat(mcp): Tasks extension (io.modelcontextprotocol/tasks) (#1283)
* feat(mcp): speak spec revision 2026-07-28 (stateless core) Adopt the 2026-07-28 MCP spec revision on the connector endpoint while keeping every handshake-era client (2025-06-18 and earlier) byte-identical: - Accept per-request _meta protocol negotiation (io.modelcontextprotocol/protocolVersion); unsupported versions return UnsupportedProtocolVersionError (-32022) with the supported list. - Implement server/discover (spec MUST): supported revisions, capabilities including the extensions field, identity, instructions, freshness hints. - Decorate results for stateless clients: required resultType, serverInfo in _meta, and CacheableResult ttlMs/cacheScope on tools/list, prompts/list, resources/list, resources/read. - Validate the standard Mcp-Method/Mcp-Name request headers when present (HeaderMismatchError -32020); absence stays accepted. - Declare the ratified MCP Apps extension (io.modelcontextprotocol/ui) in capabilities; the widgets already use the ratified mime type and _meta.ui.resourceUri shape, so no widget changes are needed. - OAuth: include the RFC 9207 iss parameter on every authorization response (success and error) and advertise authorization_response_iss_parameter_supported in RFC 8414 metadata. Resource-not-found already used -32602 and tools/list ordering was already deterministic; both are covered by the new test file. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(mcp): Tasks extension (io.modelcontextprotocol/tasks) Durable handles for long-running MCP tool calls, per the official Tasks extension. A client that declares the extension in its per-request capabilities gets a CreateTaskResult (resultType: "task") immediately; the work completes after the response via after() and lands in the new mcp_tasks table for tasks/get polling. Clients that did not declare the extension are never handed a task (spec MUST). - New mcp_tasks table (migration 20260729094000): company-scoped SELECT RLS, service-role-only writes (mirrors pending_operations), 1-hour expiry, status lifecycle CHECK. pg-real coverage included; triaged as excluded in the full-archive backup contract (transient state). - tasks/get (creator-scoped), tasks/cancel (cooperative, working-only flip), tasks/update (ack no-op: no input_required flows yet). - Tool opt-in via shouldRunAsTask predicate; first producer is gnubok_audit_package, the one genuinely long-running blocking call (multi-minute ZIP generation). estimate_only stays synchronous. - Tool failures complete the task with the standard isError envelope, exactly what the synchronous call would have returned; the failed status stays reserved for infrastructure errors. - server/discover and initialize now advertise the tasks extension. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): creator-only task RLS, enforced expiry sweep, RoPA entry Compliance-swarm follow-ups on the mcp_tasks migration (editing the migration is safe: it has not shipped beyond the ephemeral PR preview): - SELECT RLS tightened from company-wide to auth.uid() = user_id so the DB grant matches the creator-scoped tasks/get contract; task results carry raw tool output (Art. 5(1)(c)). pg test now proves a same-company colleague cannot read the row. - The 1-hour retention is now enforced, not aspirational: createMcpTask opportunistically deletes expired rows on every creation (idx_mcp_tasks_expires), best-effort (Art. 5(1)(e)). - RoPA entry mcp.async_task_handles added to .compliance/ropa.yaml (Art. 30). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): literal terminal-update payload for the phantom-column guard The conditional spreads in resolveMcpTask made the payload unresolvable for the no-phantom-columns guard (362 > 360 ceiling). A literal payload writing null for absent terminal fields is equivalent here: the terminal transition sets the complete terminal state. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
4501f118c2 |
feat(mcp): approval-queue MCP Apps widget for staged operations (#1278)
* feat(mcp): approval-queue MCP Apps widget for staged operations gnubok_list_pending_operations(render_ui=true) now renders an interactive approval queue (claude.ai / Claude Desktop) where the user approves or rejects each staged operation with a click. High-risk operations arm the approve button and the second click sends confirmed=true, so the BFL 5 kap 5 acknowledgment is a first-party human action instead of the agent asserting confirmed=true on the user's behalf (the audit weakness flagged in dev_docs/erpclaw_analysis.md). - New widget ui://pending-operations/app.html following the established self-contained postMessage/JSON-RPC pattern (no fetch, theme-aware, Swedish labels, expandable preview_data per row). - Result-level _meta.ui hint gated on render_ui=true, mirroring the VAT report wiring; the tool stays data-only by default. - Widget tool references project per namespace (accounted_* clients see accounted_ names inside the HTML). - tools/list payload ceiling 58K -> 58.5K per the in-test convention: prose trimmed to the floor first, remainder is wire contract. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): time out the widget RPC bridge so a silent host cannot strand a row Review follow-up: sendRequest never settled if the host dropped a response, leaving op._working=true forever with the approve/reject buttons gone. A 30s timeout rejects the promise; the existing catch paths restore the row with an error message so the user can retry. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
dc5aea4a35 |
feat(mcp): speak spec revision 2026-07-28 (stateless core) (#1277)
* feat(mcp): speak spec revision 2026-07-28 (stateless core) Adopt the 2026-07-28 MCP spec revision on the connector endpoint while keeping every handshake-era client (2025-06-18 and earlier) byte-identical: - Accept per-request _meta protocol negotiation (io.modelcontextprotocol/protocolVersion); unsupported versions return UnsupportedProtocolVersionError (-32022) with the supported list. - Implement server/discover (spec MUST): supported revisions, capabilities including the extensions field, identity, instructions, freshness hints. - Decorate results for stateless clients: required resultType, serverInfo in _meta, and CacheableResult ttlMs/cacheScope on tools/list, prompts/list, resources/list, resources/read. - Validate the standard Mcp-Method/Mcp-Name request headers when present (HeaderMismatchError -32020); absence stays accepted. - Declare the ratified MCP Apps extension (io.modelcontextprotocol/ui) in capabilities; the widgets already use the ratified mime type and _meta.ui.resourceUri shape, so no widget changes are needed. - OAuth: include the RFC 9207 iss parameter on every authorization response (success and error) and advertise authorization_response_iss_parameter_supported in RFC 8414 metadata. Resource-not-found already used -32602 and tools/list ordering was already deterministic; both are covered by the new test file. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(mcp): Mcp-Name covers params.uri, base64 sentinel, version-header consistency Review follow-ups against the transport spec text: Mcp-Name mirrors params.name OR params.uri (resources/read), values arrive base64-wrapped in the =?base64?...?= sentinel and must be decoded before comparison, and an MCP-Protocol-Version header that disagrees with the _meta protocol version is a HeaderMismatch. Absence of any header stays accepted since this server supports handshake-era clients (spec-sanctioned leniency). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob Wennberg
|
80a14ddfd2 |
feat(mcp): dimension parity for the write/read tool edges (#1274)
Closes the MCP dimension gaps found in the 2026-07-28 audit:
- gnubok_bulk_book_inbox_items accepts a shared dimensions bag through
all three layers (tool schema + BulkBookInboxSchema + categorize-core
BulkBookInboxInput), resolve-don't-select with echoed resolutions; the
web inbox bulk-book route and the pending-op executor inherit it via
the shared schema.
- gnubok_create_employee / gnubok_update_employee accept
default_dimensions (names resolve to codes; {} clears on update).
The command layer already persisted the field: only the MCP boundary
blocked it, leaving payroll tagging dashboard-only.
- gnubok_query_journal: dimensions bag filter (jsonb containment via
the GIN index, covers custom dims the legacy project/cost_center
filters cannot) + include_dimensions to return each line's bag.
The wide full-match fetch stays dims-free unless something needs it.
- gnubok_list_invoices / gnubok_list_supplier_invoices return
default_dimensions (agents could set invoice bags but never read
them back).
- Discoverability: create_voucher, categorize_transaction,
correct_entry, update_invoice descriptions now name dimensions;
categorize_month and invoice_run loadouts include
gnubok_list_dimensions. Trimmed new schema prose to stay under the
tools/list payload budget.
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|