70e893b8d4c2d5348e7c6f7ab7407fa1b2eabbed
6
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Jakob WennbergandClaude Fable 5
|
ea236cbcdf |
fix(reconciliation): Bankavstämning phase 0 — correctness + feedback batch (+ nav IA regrouping) (#879)
* feat(nav): interaction-mode sidebar grouping — Arbeta/Analys/Data/Skatt & bokslut Nav IA redesign phase 0 (dev_docs/nav_ia_redesign.md): same routes, regrouped by what the user is doing. CLAUDE.md restructured around Hard Rules (doc references updated); pending-page explainer removed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(reconciliation): correctness + feedback batch for Bankavstämning (phase 0) Engine: fetchAllRows pagination on status/run/RPC fetches (silent 1000-row cap corrupted totals), optimistic-lock guards on manualLink + apply, unlink audit rows attributed to the acting user (was: company UUID), selected_matches partial apply intersected with a fresh match run. View: silent in-place refresh instead of a full-page skeleton per action, checkbox-gated apply with confidence badges (fuzzy unticked) in chunks of 500, honest result toasts, dry-run errors surfaced, ranked per-row picker candidates pinned to the applied date window, currency-correct amounts (bank side in account currency, GL side SEK), voucher links, translated source types, colored differens, dirty-date-filter guard. Discovery: year-end preflight 404 href fixed (/reconciliation/bank never existed), ⌘K palette entry, real links from the transactions page. v1: status registry schema now matches the actual ReconciliationStatus payload, errors documented as a count, false ~0.85-threshold pitfall replaced, route test mocks the real shape. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob WennbergandClaude Fable 5
|
59ecaee650 |
feat(mcp): qualified identifiers in all tool output schemas (P1-2) (#877)
Agents grabbed the wrong id when list rows exposed a bare 'id' next to qualified ids (journal_entry_id) with no type distinction, got NOT_FOUND, and had to re-derive (agent.feedback). Additive sweep of all 11 bare-id sites: every output identifier now ships a fully qualified sibling (transaction_id, fact_id, atom_id, company_id, dimension_id, dimension_value_id); the bare 'id' stays as a deprecated alias for compatibility. New ratchet test (qualified-ids.test.ts): a shrinking grandfathered list carries the deprecated aliases; any NEW bare 'id' in an output schema fails CI, every remaining alias must ship alongside its qualified sibling, and stale grandfather entries must be pruned. Convention documented in .claude/rules/mcp-server.md. Full unit suite: 6,577 green. Part of dev_docs/mcp_optimization_plan.md (P1-2). Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Jakob WennbergandClaude Fable 5
|
250cc7c450 |
feat(mcp): always-explicit retryable on structured errors + transient inference (P1-1) (#875)
Agents could not distinguish 'keep retrying' from 'stop, this is
broken' (agent.feedback): retryable was emitted only when a registry
entry declared true — absent otherwise, including for genuinely
transient DB/network failures whose SQLSTATE is lost when tools wrap
them as Error('Database error: ...').
- StructuredError.retryable is now a required boolean. Registry
declaration wins; otherwise isTransientFailure() infers from Postgres
SQLSTATEs (40001/40P01/57014/08xxx/53xxx/55P03), upstream HTTP
statuses (408/429/5xx), and message signatures that survive wrapping
(deadlock, serialization, statement timeout, fetch/socket failures).
- Unclassified transient failures surface as stable code
TRANSIENT_ERROR (new registry entry, retryable: true).
- categorize_transaction accepts idempotency_key — the tool agents
blind-retry after client-side approval-elicitation drops; the key
makes that retry replay-safe instead of double-staging.
- Contract documented in .claude/rules/mcp-server.md. The planned
'kind' field was dropped: the code registry already encodes it;
retryable is the agent-actionable bit.
Every tool error already flows through the single dispatch point
(toToolError -> getStructuredError), so coverage is universal without
per-tool migration. Full unit suite: 6575 tests green.
Part of dev_docs/mcp_optimization_plan.md (P1-1).
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|
Jakob WennbergandClaude Opus 4.8
|
43007fa869 |
feat(mcp): self-describing agent surface — staging _meta, company identity, clean skill summaries (#775)
* feat(mcp): make the agent surface self-describing (staging _meta, company identity, clean summaries)
A pass over the MCP server's agent-facing surface so an agent can act
correctly without parsing description prose:
- Machine-readable staging contract: deriveToolMeta() attaches _meta to
tools/list (and search detail=full) — { requires_approval, approve_tool,
preflight? } — keyed off the STAGED_OPERATION_SCHEMA output schema. Literal
_meta (e.g. UI widget hints) wins on collision. TOOL_PREFLIGHT_MAP names the
read-only pre-flight for the few writes that have one (year-end readiness,
VAT validate, depreciation proposal). Guarded by staging-meta.test.ts.
- Company identity in gnubok_get_agent_briefing: returns a `company` block
(id, name, org_number, entity_type, accounting_method) so the agent can
confirm WHICH entity it operates on and pick the right settlement account
(accrual = credit 1510; cash = debit 19xx) before any write. Best-effort —
a missing row never blocks the briefing. Covered by agent-briefing.test.ts.
- toSummary(): trims the long, keyword-stuffed SKILL.md frontmatter into clean
one-liners for gnubok_list_skills / gnubok_get_agent_briefing so the client
never truncates one mid-sentence; full bodies stay in gnubok_load_skill.
Covered by to-summary.test.ts.
- bank-reconciliation skill: a match/link decision tree (what you have x
whether a verifikat exists) and kontant- vs faktureringsmetoden settlement
accounts.
- Prose/description clarifications: "Stages"/"Stages for approval" on the
link tools; propose_dispositioner/accruals note there is no dedicated MCP
poster; server-info documents _meta and the legacy gnubok_ tool prefix.
All 34 touched MCP tests pass. Merged cleanly on top of #759/#760 (server.ts).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(mcp): make accounting_method description state the full settlement posting
Review (swedish-accounting-compliance): the agent-briefing schema described
accrual as "credit 1510 on payment", which reads as a one-sided entry. Spell
out both sides (payment debits 19xx AND credits 1510) so an agent can't infer a
single-leg posting that violates BFL 5 kap double-entry. Mirrors the precision
already in the bank-reconciliation skill body. Payload-size guard still passes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
||
|
Jakob WennbergandClaude Opus 4.8
|
0b86901a2b |
Enforce MFA on critical mutation routes + post-audit foundation (A1) (#646)
* feat(lib): add canonical money + format + fetch primitives (audit Tier 0) Foundation for post-audit cleanup: shared primitives so subsequent refactors import one helper instead of reinventing (the duplication the audit found). - lib/money.ts: canonical roundOre/ORE_TOLERANCE (+ equalOre/isZeroOre/sumOre); lib/bokslut/rounding.ts re-exports for back-compat - lib/utils.ts: formatAmount, formatWholeKr, formatDateTime - lib/hooks/use-fetch.ts: generic client fetch hook (abort, bilingual errors, refetch) - components/common/DataState.tsx: loading/error/empty wrapper over Skeleton/EmptyState - messages: common.retry / common.load_error (sv+en) - tests: 16 tests incl. the 1.005 half-ore case and locale-robust format assertions Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * ci(guards): ratchet against new MFA-bypassing routes and naive ore-rounding Adds scripts/checks/no-new-antipatterns.mjs + committed baseline. Fails CI only when a PR ADDS a route hand-rolling supabase.auth.getUser() (which skips MFA AAL2 enforcement) or a new Math.round(x*100)/100. Baseline: 178 raw-auth routes, 668 naive rounds — ratchets down as the A1 (route-auth) and D1 (rounding) migrations land. Wired into core-build.yml; green at baseline. Note: scripts/ is gitignored (.gitignore:70 '/scripts') yet tracks 39 files via force-add; these two were force-added to match that existing pattern. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(api,errors): enforce MFA on journal-entry mutation routes via withRouteContext (A1) Migrates the 4 journal-entry mutation routes (commit, correct, reverse, recordate) off hand-rolled supabase.auth.getUser() onto withRouteContext, which enforces MFA AAL2 (requireAuth) + non-viewer role (requireWrite) and routes thrown errors through the canonical errorResponse envelope. Fixes audit finding A1 for the most compliance-critical mutations and folds in C8 for these routes (drops bookkeepingErrorResponse; they now emit message_en). Also fixes a latent bug: errorResponse()/extractBookkeepingDetails only handled 11 of 15 typed bookkeeping errors, so MeaninglessCorrection / NoOpenPeriodForDate / TargetPeriodClosed / TargetPeriodLocked silently degraded to a generic 500 (affecting existing v1 callers too). Adds the 4 missing registry codes + extract cases -> correct 400/409. Behavior change: untyped engine throws now return the canonical 500 envelope instead of 400+raw-string; typed errors keep their status (verified against the registry). Tests updated to the realistic typed-error contract + a 403 write-gate test on commit. Updates .claude/rules/api-routes.md to prescribe withRouteContext. Ratchets the antipattern guard 178 -> 174. Full unit suite green (5023); tsc: no new errors. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * feat(api): enforce MFA on salary run authorization routes via withRouteContext (A1) Migrates the salary-run lifecycle write routes (approve, paid, revert) — the highest-PII A1 surface — off hand-rolled supabase.auth.getUser() onto withRouteContext (enforces MFA AAL2 + non-viewer role). Explicit { error } returns are preserved unchanged (passed through the wrapper); only auth changes, so no error-shape regression. Salary unit suite green (8). Ratchets the antipattern guard 174 -> 171. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * review: address PR #646 bot findings - guard: match withRouteContext/requireAuth at the CALL site (withRouteContext[<(]), not a bare import — closes the false-negative greptile flagged. It surfaced app/api/sandbox/seed (hand-rolled getUser; the loose regex had matched a code comment). Switched that route to requireAuth() — the documented stopgap for routes that can't use withRouteContext (it runs before a company exists; anonymous users, so MFA is a no-op but the auth path is now consistent). Guard stays at 171. - money.test: add the negative half-ore case roundOre(-1.005) === -1 to lock the rounding direction against regressions. - use-fetch: document keep-previous-data + deferred-loading (effect-tick) semantics. - structured-errors: drop the BFL 5 kap. 5 § citation from MEANINGLESS_CORRECTION per the swedish-compliance bot (5 § governs correction procedure, not the no-op precondition). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> * review: enrich wrapper error logging + document sandbox GDPR controls (PR #646) - with-route-context: log unhandled errors and route errorResponse through the resolved { userId, companyId } logger, not just { requestId, operation } — closes the OWASP V16 audit-trail finding for all 82+ routes using the wrapper. Documented in the JSDoc. - sandbox/seed: document the GDPR Art.32 compensating controls for the anonymous write path (anonymous-only, /24 rate limit, synthetic demo data, own-company RLS scope). No functional change — the flagged behaviour is pre-existing by design; this records the reasoning inline. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
Jakob WennbergandClaude Opus 4.8
|
c74b19df1b |
Accounted rebrand + swarm-skill cleanup + bank-reconciliation fixes (#643)
* feat(reconciliation): close the bank-feed loop on voucher links and re-tag mis-typed opening balances
Two related fixes to bank reconciliation correctness:
1. Auto-reconcile on voucher link. Linking an invoice or supplier invoice to
an existing voucher previously advanced only the invoice — the bank
transaction that paid it kept sitting in the Transactions inbox with a null
journal_entry_id. linkInvoiceToVoucher / linkSupplierInvoiceToVoucher now
call autoReconcileTransactionForLinkedVoucher (lib/reconciliation), which
links the bank transaction to the same verifikat when exactly one unbooked
line matches it. Best-effort and post-commit: a failure here never fails the
link. The result surfaces reconciledTransactionId; the inbox row leaves the
list and the UI shows link_success_tx_reconciled.
2. Re-tag mis-typed opening balances. getReconciliationStatus and the GL-line
matching RPCs identify a cash account's ingående balans solely by
journal_entries.source_type='opening_balance'. Companies migrated from other
systems often booked the bank IB as an ordinary voucher (source_type
'import' or 'manual'), so it was never excluded and surfaced as a phantom
reconciliation difference equal to the opening balance. Adds:
- migration mark_entry_as_opening_balance: a GUC-gated carve-out in the
immutability trigger plus a SECURITY DEFINER RPC that validates the entry
(balance-sheet lines only, dated on a fiscal-period boundary), flips the
source_type, and writes an audit row — no blanket data sweep.
- POST /api/reconciliation/bank/mark-opening-balance + MarkOpeningBalanceSchema.
- BankReconciliationView action to trigger it from the IB diff.
The gnubok_create_voucher executor now accepts a typed is_opening_balance flag
and derives source_type='opening_balance' only after validating class 1/2 lines
on the period start, so new IBs land correctly typed.
Covered by lib/reconciliation auto-reconcile tests, voucher-executors tests,
and a mark-entry-as-opening-balance pg-real test.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* chore: rebrand gnubok → Accounted and prune swarm agent skills
Product rebrand and skills housekeeping. No runtime behaviour change.
Rebrand: replace user-visible "gnubok" with "Accounted" across docs, READMEs,
in-code comments, doc-site content, MCP skill/resource prose, and the
gnubok-mcp package description. The MCP resource URI scheme is moved gnubok://
→ Accounted:// consistently across resource registrations, the event-type
comment, and the resource/skill tests. Deliberately preserved as stable
identifiers (NOT rebranded): the gnubok-company-id cookie, gnubok_sk_ / gnubok_inv_
token prefixes, the gnubok-mcp npm bridge name, and the AGI <gem:Programnamn>
value (kept 'gnubok' per its source comment — it is the software identifier sent
to Skatteverket and must not churn across visual rebrands).
Skills: remove the 27 swarm-* agent SKILL.md atoms (no longer used; already
absent from the agent_atom_registry in prod), refresh the remaining skill docs,
add the .claude/rules/ path-scoped rule set, and regenerate the
seed_agent_atom_bodies migration + .skill-body-manifest.json via
`npm run skills:generate` so the DB-backed skill bodies match the trimmed set.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|