fix(errors): keep the SQLSTATE when wrapping database errors (#2027)

isTransientFailure() checks the driver's error code first, and 57014
(statement timeout) is already in its transient set. But the wrapping idiom
across the codebase was `throw new Error(\`Database error: ${err.message}\`)`,
which keeps the prose and drops the code. A retryable timeout therefore
arrived anonymous and resolved to UNKNOWN_ERROR: "Något gick fel. Försök
igen." An agent cannot dispatch on that, so it retried.

On production over 60 days, with the two bot integrations excluded: 1024 real
agent failures, 645 of them UNKNOWN_ERROR across 60 actors and 57 companies.
82 retry streaks of three or more identical failures, 462 wasted repeat calls,
53.1% of all agent error calls sitting inside a streak.

The worst offender traces to one line in core. gnubok_query_journal failed 164
times at a p50 of 8110ms while every other failing tool sat between 1 and
315ms, and its path is fetchEntryLines -> fetchAllRows, where
lib/supabase/fetch-all.ts threw `new Error(error.message)`. That is the
highest-traffic strip point in the repo: 31 callers, every paginated read.
query_journal already had a correct TRANSIENT_ERROR branch offering "retry, or
narrow with date_from/date_to" which could never fire, because by the time it
looked, the code was gone.

fetch-all keeps the driver message verbatim: callers match on the existing
text, and this adds the code rather than rewording anything.

Attaching the code is safe. extractCode() only accepts /^[A-Z_]+$/ and every
SQLSTATE contains digits, so it cannot be mistaken for one of our own stable
codes. There is a test for that, and one asserting the old bare-Error shape
still resolves to UNKNOWN_ERROR so the fix cannot silently regress.

Also stops rendering the literal "undefined" when a driver-level failure
carries no message, which is the string that made these unsearchable.

Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Jakob Wennberg
2026-08-30 10:48:05 +02:00
committed by GitHub
parent 7e76961da1
commit 4b7343d5ec
6 changed files with 302 additions and 55 deletions
+1
View File
@@ -1337,3 +1337,4 @@ One line per decision: `[YYYY-MM-DD] <decision>: <why>`. Appended by agents and
[2026-08-28] /migrate SIE guard skips company-info-only runs (all entity flags false) and the wizard derives "SIE already imported" from the preview OR this session's successful /import-sie results: company info writes no accounts, balances or subledger rows, so the BFL rationale does not apply; and the one-shot preview went stale after phase 1 succeeded and phase 2 failed, falsely blocking an entities-only retry (#2000 review).
[2026-08-28] get_vat_ruta_source_lines (the VAT ruta drill-down) now applies the same four exclusions as get_vat_declaration_totals (the filed figure): posted closing entries, source_type 'vat_settlement', the two kontantmetod year-end reversals, and settlement-SHAPED entries (a line on a ruta account plus a line on 2650/1650). It previously filtered on company, status and date only, so expanding a ruta listed verifikat that are not in the number it claims to explain, with no total on the panel to reveal the mismatch. Measured on prod 2026-08-28: 322 posted/reversed entries carrying 26xx lines across 214 companies sit in those excluded classes. A momsdeklaration is räkenskapsinformation (BFL 5 kap.) and this drill-down is what substantiates a filed figure, so the two must agree exactly. The exclusion CTEs are lifted VERBATIM from the figure rather than re-derived: any divergence reintroduces exactly this bug, and an identical copy is easy to diff when the figure changes. Settlement-shape is detected against journal_entry_lines directly instead of through the figure's vat_lines CTE, which is EQUIVALENT not a shortcut (p_ruta_accounts = VAT_ACCOUNTS and p_net_accounts = ['2650','1650'] are both strict subsets of the figure's p_accounts, so restricting to vat_lines first cannot change which entries match); that keeps p_accounts meaning "the accounts of the ruta being expanded" without a fourth account parameter. opening_balance entries are deliberately NOT excluded: the figure exempts them from `shaped`, which keeps their lines IN the totals, so dropping them here would break the equality in the other direction (pinned by its own test). VAT_ACCOUNTS is now exported from lib/reports/vat-declaration.ts so the route detects shape from the same list the figure uses; a second copy is what let the two disagree. DROP + CREATE OR REPLACE, not CREATE OR REPLACE alone: the signature gains p_ruta_accounts/p_net_accounts and adding parameters registers a second overload PostgREST cannot choose between (trap documented in 20260421140000); OR REPLACE on the new arity keeps the file re-runnable. Verified the new pg test actually catches the bug by reinstalling the old body and watching 3 of 4 tests fail with the real misreporting (2611: drill-down 250/240 vs figure 0/200), then restoring.
[2026-08-28] Bankavstamning NULL-link fix scoped to transfer legs with contradicting sign (20260828220000): the naive rule (NULL counts only for the primary account) and the formula-only variant (drop far-leg-settled vouchers from unexplained) were both simulated against prod and rejected; the naive rule worsened 4 of 11 affected cards (worst -37 000 kr false alarm on single-leg vouchers with no user action available), the formula variant blew up healthy cards by up to 474 550 kr. The shipped three-condition rule changes 24 vouchers on 7 cards in 6 companies, all verified per-card.
[2026-08-29] Database errors now keep their SQLSTATE: new lib/errors/db-error.ts (dbError/errorCauseTag), applied at the 54 `throw new Error(\`Database error: ${err.message}\`)` sites in the MCP server AND, far more importantly, at lib/supabase/fetch-all.ts:74 where `throw new Error(error.message)` was the single highest-traffic strip point in the codebase (31 callers; every paginated read). isTransientFailure() checks the driver code FIRST and 57014 (statement timeout) is already in TRANSIENT_SQLSTATES, so discarding it turned a retryable timeout into UNKNOWN_ERROR ("Något gick fel. Försök igen."), which an agent cannot dispatch on. Traced end to end: gnubok_query_journal -> fetchEntryLines -> fetchAllRows (code stripped here) -> the tool's own sanitizeDbError, which ALREADY had a correct TRANSIENT_ERROR branch with a "retry or narrow with date_from/date_to" hint that could never fire because getStructuredError saw an anonymous Error. Measured on prod over 60 days with bot actors excluded: 1 024 real-agent failures, 645 UNKNOWN_ERROR across 60 actors and 57 companies; query_journal failed 164 times at p50 8 110 ms while every other failing tool sat at 1-315 ms; 82 retry streaks, 462 wasted repeat calls, 53.1% of error calls inside a streak. fetch-all passes context=null so the driver message stays VERBATIM (sanitizeDbError and other callers match on the existing text; this change adds the code, it does not reword). Attaching `code` is safe because extractCode() only accepts /^[A-Z_]+$/ and every SQLSTATE/PostgREST code contains digits, so it cannot hijack the application error registry (pinned by a test). dbError also never renders the literal "undefined": a driver-level failure with no message produced "Database error: undefined", the string that made these unsearchable. errorCauseTag() returns a PII-safe SQLSTATE for telemetry; the raw driver message can quote row values in a constraint violation and belongs in the server log, never in event_log. NOT ratcheted: check:types reports 538 vs baseline 539 because main fixed an unrelated error in own-account-detector.test.ts after the baseline was set; the gate only fails on an INCREASE, so the baseline is left alone rather than adding unrelated churn to this diff.
+55 -54
View File
@@ -47,6 +47,7 @@ import { applyAccountOverride } from '@/lib/bookkeeping/account-override'
import { ACCOUNT_NUMBER_RE } from '@/lib/invariants/account-number'
import { isSlpPensionAccount } from '@/lib/bookkeeping/slp-lines'
import { getErrorEntry } from '@/lib/errors/structured-errors'
import { dbError } from '@/lib/errors/db-error'
import { getStructuredError } from '@/lib/errors/get-structured-error'
import { applySettlementAccount } from '@/lib/bookkeeping/mapping-engine'
import { resolveSettlementAccount } from '@/lib/bookkeeping/settlement-account'
@@ -1055,7 +1056,7 @@ async function resolveJournalEntryRef(
.order('entry_date', { ascending: false })
if (error) {
throw new Error(`Database error resolving voucher "${series}-${number}": ${error.message}`)
throw dbError(error, `Database error resolving voucher "${series}-${number}"`)
}
const matches = (data ?? []) as Array<{ id: string; entry_date: string; description: string }>
@@ -3799,7 +3800,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!data) throw new Error('Company settings not found.')
return {
@@ -3916,7 +3917,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!current) throw new Error('Company settings not found.')
const currentPreview = {
@@ -5041,7 +5042,7 @@ export const tools: McpTool[] = [
if (cashAccountId) countQuery = countQuery.eq('cash_account_id', cashAccountId)
const { count: totalCount, error: countError } = await countQuery
if (countError) throw new Error(`Database error: ${countError.message}`)
if (countError) throw dbError(countError)
let listQuery = supabase
.from('transactions')
@@ -5055,7 +5056,7 @@ export const tools: McpTool[] = [
.order('date', { ascending: false })
.range(offset, offset + limit - 1)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
// Resolve the bank account's BAS ledger for the rows on this page so a
// per-account reconciliation can be driven from outside (customer
@@ -5070,7 +5071,7 @@ export const tools: McpTool[] = [
.select('id, ledger_account')
.eq('company_id', companyId)
.in('id', cashAccountIds)
if (cashError) throw new Error(`Database error: ${cashError.message}`)
if (cashError) throw dbError(cashError)
for (const c of (cashRows ?? []) as Array<{ id: string; ledger_account: string }>) {
ledgerByCashAccount.set(c.id, c.ledger_account)
}
@@ -5150,7 +5151,7 @@ export const tools: McpTool[] = [
p_limit: limit,
p_offset: offset,
})
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const result = data as {
ok: boolean
@@ -5228,7 +5229,7 @@ export const tools: McpTool[] = [
p_limit: limit,
p_offset: offset,
})
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const result = data as {
ok: boolean
@@ -5490,7 +5491,7 @@ export const tools: McpTool[] = [
.order('date', { ascending: false })
.limit(limit)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
return {
transactions: data ?? [],
@@ -5551,7 +5552,7 @@ export const tools: McpTool[] = [
.range(from, to)
})
} catch (error) {
throw new Error(`Database error: ${error instanceof Error ? error.message : 'unknown error'}`)
throw dbError(error)
}
rows.sort((a, b) => a.name.localeCompare(b.name, 'sv') || a.id.localeCompare(b.id))
@@ -5843,7 +5844,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!current) throw new Error('Customer not found.')
// Same guard as gnubok_create_customer and the REST PATCH route: only
@@ -5971,7 +5972,7 @@ export const tools: McpTool[] = [
return q.order('id', { ascending: true }).range(from, to)
})
} catch (error) {
throw new Error(`Database error: ${error instanceof Error ? error.message : 'unknown error'}`)
throw dbError(error)
}
articles.sort((a, b) => a.name.localeCompare(b.name, 'sv') || a.id.localeCompare(b.id))
@@ -6157,7 +6158,7 @@ export const tools: McpTool[] = [
.order('id', { ascending: false })
.range(offset, offset + limit)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const rows = data ?? []
const invoices = rows.slice(0, limit).map((inv: Record<string, unknown>) => ({
@@ -6285,7 +6286,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!invoice) throw new Error('Invoice not found. Use gnubok_list_invoices to find valid IDs.')
type InvoiceLineRow = {
@@ -7258,7 +7259,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (invoiceError) throw new Error(`Database error: ${invoiceError.message}`)
if (invoiceError) throw dbError(invoiceError)
if (!invoice) throw new Error('Invoice not found')
// Never read invoice_deliveries directly: the row carries the exact
@@ -7270,7 +7271,7 @@ export const tools: McpTool[] = [
'list_invoice_delivery_summaries_for_service',
{ p_company_id: companyId, p_user_id: userId, p_invoice_id: invoiceId },
)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
// Mirrors the RETURNS TABLE of the RPC. body_html, body_text and
// bcc_addresses are absent by construction, not filtered here.
@@ -7411,7 +7412,7 @@ export const tools: McpTool[] = [
.range(from, to)
})
} catch (error) {
throw new Error(`Database error: ${error instanceof Error ? error.message : 'unknown error'}`)
throw dbError(error)
}
suppliers.sort((a, b) => a.name.localeCompare(b.name, 'sv') || a.id.localeCompare(b.id))
@@ -7605,7 +7606,7 @@ export const tools: McpTool[] = [
const { data, error } = await query.order('due_date', { ascending: true }).limit(limit)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
return { invoices: data ?? [], count: data?.length ?? 0 }
},
@@ -7650,7 +7651,7 @@ export const tools: McpTool[] = [
.order('occurrence_count', { ascending: false })
.limit(limit)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
return {
templates: (data ?? []).map((t) => ({
@@ -7710,7 +7711,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.in('id', limitedIds)
if (txError) throw new Error(`Database error: ${txError.message}`)
if (txError) throw dbError(txError)
if (!transactions || transactions.length === 0) throw new Error('No transactions found')
// Fetch mapping rules
@@ -7846,7 +7847,7 @@ export const tools: McpTool[] = [
return query.order('account_number', { ascending: true }).range(from, to)
})
} catch (error) {
throw new Error(`Database error: ${error instanceof Error ? error.message : 'unknown error'}`)
throw dbError(error)
}
// Postgres ordered by sort_order ascending with nulls last; keep that
@@ -7917,7 +7918,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.eq('account_number', accountNumber)
.maybeSingle()
if (existingErr) throw new Error(`Database error: ${existingErr.message}`)
if (existingErr) throw dbError(existingErr)
if (existing) {
throw new Error(
existing.is_active
@@ -8038,7 +8039,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.eq('account_number', accountNumber)
.maybeSingle()
if (fetchErr) throw new Error(`Database error: ${fetchErr.message}`)
if (fetchErr) throw dbError(fetchErr)
if (!current) {
throw new Error(`Konto ${accountNumber} finns inte i kontoplanen. Skapa det med gnubok_create_account.`)
}
@@ -8240,7 +8241,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.eq('sie_dim_no', sieDimNo)
.maybeSingle()
if (dimError) throw new Error(`Database error: ${dimError.message}`)
if (dimError) throw dbError(dimError)
if (!dimension) {
throw new Error(
`Dimension ${sieDimNo} finns inte i registret. Anropa gnubok_list_dimensions för att se registrerade dimensioner.`,
@@ -8256,7 +8257,7 @@ export const tools: McpTool[] = [
if (!includeInactive) valuesQuery = valuesQuery.eq('is_active', true)
const { data: rows, error: valuesError } = await valuesQuery
if (valuesError) throw new Error(`Database error: ${valuesError.message}`)
if (valuesError) throw dbError(valuesError)
const all = (rows ?? []) as Array<{
id: string
code: string
@@ -8350,7 +8351,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.eq('sie_dim_no', params.sie_dim_no)
.maybeSingle()
if (dimError) throw new Error(`Database error: ${dimError.message}`)
if (dimError) throw dbError(dimError)
if (!dimension) {
throw new Error(
`Okänd dimension ${params.sie_dim_no}. Endast registrerade dimensioner kan få nya värden: ` +
@@ -8370,7 +8371,7 @@ export const tools: McpTool[] = [
.eq('dimension_id', dimension.id)
.eq('code', params.code)
.maybeSingle()
if (existingError) throw new Error(`Database error: ${existingError.message}`)
if (existingError) throw dbError(existingError)
if (existing?.is_active) {
throw new Error(
`Värdet "${params.code}" (${existing.name}) finns redan i ${dimension.name}: använd koden direkt i dimensions-baggen.`,
@@ -10925,7 +10926,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.order('period_start', { ascending: false })
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const periods = (data ?? []).map((p) => ({
id: p.id,
@@ -11719,7 +11720,7 @@ export const tools: McpTool[] = [
}
const { data, error } = await query
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const mapped = (data || []).map((item) => {
const extracted = item.extracted_data as Record<string, unknown> | null
@@ -11819,7 +11820,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.single()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!data) throw new Error('Inbox item not found')
return data
@@ -12275,7 +12276,7 @@ export const tools: McpTool[] = [
}
const { data: inboxRows, error: inboxError } = await inboxQuery
if (inboxError) throw new Error(`Database error: ${inboxError.message}`)
if (inboxError) throw dbError(inboxError)
if (!inboxRows || inboxRows.length === 0) {
return { items: [], count: 0 }
}
@@ -12287,7 +12288,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.in('document_id', docIds)
if (txError) throw new Error(`Database error: ${txError.message}`)
if (txError) throw dbError(txError)
const matchedDocIds = new Set((txMatches || []).map((t) => t.document_id))
const unmatched = inboxRows
@@ -12402,7 +12403,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (docError) throw new Error(`Database error: ${docError.message}`)
if (docError) throw dbError(docError)
if (!doc) throw new Error('Document not found')
const ttlSeconds = 300
@@ -12717,7 +12718,7 @@ export const tools: McpTool[] = [
.from('fiscal_periods')
.select('id, period_start, period_end')
.eq('company_id', companyId)
if (periodsError) throw new Error(`Database error: ${periodsError.message}`)
if (periodsError) throw dbError(periodsError)
const periodsByYear = new Map<number, Array<{ id: string; period_start: string; period_end: string }>>()
for (const p of periods ?? []) {
@@ -12734,7 +12735,7 @@ export const tools: McpTool[] = [
.select('id, file_name, mime_type, journal_entry_id')
.in('id', documentIds)
.eq('company_id', companyId)
if (docsError) throw new Error(`Database error: ${docsError.message}`)
if (docsError) throw dbError(docsError)
const docsById = new Map((docs ?? []).map((d) => [d.id as string, d as {
id: string; file_name: string; mime_type: string; journal_entry_id: string | null
}]))
@@ -12820,7 +12821,7 @@ export const tools: McpTool[] = [
.select('id, status')
.eq('company_id', companyId)
.in('id', existingJeIds)
if (existingErr) throw new Error(`Database error: ${existingErr.message}`)
if (existingErr) throw dbError(existingErr)
for (const je of existingJes ?? []) {
if ((je as { status: string }).status === 'posted') postedExistingJeIds.add(je.id as string)
}
@@ -13185,7 +13186,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
if (activeOnly) query = query.eq('is_active', true)
const { data, error } = await query.order('last_name')
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
// Shared masking helper: strips personnummer (ciphertext) AND
// personnummer_last4, exposing only personnummer_masked: same shape as
// the app routes and the other MCP payroll tools.
@@ -13370,7 +13371,7 @@ export const tools: McpTool[] = [
.eq('id', id)
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!run) throw new Error('Salary run not found')
if (run.status === 'booked') throw new Error('Salary run is already booked')
if (!['draft', 'review', 'approved', 'paid'].includes(run.status as string)) {
@@ -13383,7 +13384,7 @@ export const tools: McpTool[] = [
.from('salary_run_employees')
.select('id, calculation_breakdown')
.eq('salary_run_id', id)
if (rosterError) throw new Error(`Database error: ${rosterError.message}`)
if (rosterError) throw dbError(rosterError)
const rosterRows = roster ?? []
const uncalculated = rosterRows.filter((r) => !r.calculation_breakdown).length
if (uncalculated > 0) {
@@ -13880,7 +13881,7 @@ export const tools: McpTool[] = [
.eq('id', employeeId)
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!e) throw new Error('Employee not found')
// LLM context is a leak surface: personnummer is ALWAYS masked on MCP,
// there is no full-value drill-in on this surface.
@@ -13977,7 +13978,7 @@ export const tools: McpTool[] = [
.eq('employee_id', employeeId)
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!sre) throw new Error('Employee not found in this salary run')
const emp = sre.employee as { first_name: string; last_name: string; personnummer: string } | null
const lineItems = ((sre.line_items ?? []) as Array<Record<string, unknown>>)
@@ -14593,7 +14594,7 @@ export const tools: McpTool[] = [
.eq('id', employee_id)
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!existing) throw new Error('Employee not found')
const changes = Object.entries(patch).map(([field, to]) => ({
@@ -14714,7 +14715,7 @@ export const tools: McpTool[] = [
.select(`employee_id, ${MERGEABLE_FIELDS.join(', ')}`)
.eq('company_id', companyId)
.in('employee_id', employeeIds)
if (storedErr) throw new Error(`Database error: ${storedErr.message}`)
if (storedErr) throw dbError(storedErr)
const storedByEmployee = new Map(
((storedRows ?? []) as unknown as Array<Record<string, unknown>>).map((r) => [
r.employee_id as string,
@@ -14837,7 +14838,7 @@ export const tools: McpTool[] = [
.order('vacation_year_start', { ascending: false })
.limit(1)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!balance) throw new Error('No vacation balance exists for the employee yet (the ledger seeds on first booking)')
const { id, ...rest } = balance as { id: string } & Record<string, unknown>
const entitled = (rest.entitled_days as number) ?? 0
@@ -14855,7 +14856,7 @@ export const tools: McpTool[] = [
.eq('id', employeeId)
.eq('company_id', companyId)
.maybeSingle()
if (empErr) throw new Error(`Database error: ${empErr.message}`)
if (empErr) throw dbError(empErr)
if (!employee) throw new Error(`Employee ${employeeId} not found for this company`)
const savedTotal = Object.values((rest.saved_days as Record<string, number> | null) ?? {})
.reduce((s, d) => s + (Number(d) || 0), 0)
@@ -16121,7 +16122,7 @@ export const tools: McpTool[] = [
.eq('id', journalEntryId)
.eq('company_id', companyId)
.maybeSingle()
if (fetchErr) throw new Error(`Database error: ${fetchErr.message}`)
if (fetchErr) throw dbError(fetchErr)
if (!entry) throw new Error('Verifikationen hittades inte.')
const voucherLabel = entry.voucher_number
@@ -16555,7 +16556,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!invoice) throw new Error('Invoice not found. Use gnubok_list_invoices to find valid IDs.')
// Editable drafts only: the shared predicate the web PATCH route gates
@@ -16700,7 +16701,7 @@ export const tools: McpTool[] = [
.select('line_type, description, quantity, unit, unit_price, line_total, vat_rate, revenue_account, article_id, deduction_type, accrual_period_start, accrual_period_end')
.eq('invoice_id', invoice.id)
.order('sort_order', { ascending: true })
if (currentError) throw new Error(`Database error: ${currentError.message}`)
if (currentError) throw dbError(currentError)
currentItems = (currentRows ?? []).map((row: Record<string, unknown>) => ({
line_type: row.line_type ?? 'product',
description: row.description,
@@ -17715,7 +17716,7 @@ export const tools: McpTool[] = [
const original = data as OriginalRow | null
if (origErr) {
throw new Error(`Database error looking up journal entry ${entryId}: ${origErr.message}`)
throw dbError(origErr, `Database error looking up journal entry ${entryId}`)
}
if (!original) {
throw new Error(
@@ -17883,7 +17884,7 @@ export const tools: McpTool[] = [
const original = data as OriginalRow | null
if (origErr) {
throw new Error(`Database error looking up journal entry ${entryId}: ${origErr.message}`)
throw dbError(origErr, `Database error looking up journal entry ${entryId}`)
}
if (!original) {
throw new Error(
@@ -18897,7 +18898,7 @@ export const tools: McpTool[] = [
.order('id', { ascending: false })
.range(offset, offset + limit)
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
const rows = data ?? []
const schedules = rows.slice(0, limit).map((row: Record<string, unknown>) => {
@@ -19095,7 +19096,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!customer) throw new Error('Customer not found. Use gnubok_list_customers to find IDs.')
if (params.auto_send && !customer.email) {
throw new Error('Customer has no email address: auto_send requires one. Stage with auto_send=false or add an email first.')
@@ -19290,7 +19291,7 @@ export const tools: McpTool[] = [
.eq('company_id', companyId)
.maybeSingle()
if (error) throw new Error(`Database error: ${error.message}`)
if (error) throw dbError(error)
if (!current) throw new Error('Recurring schedule not found. Use gnubok_list_recurring_schedules to find IDs.')
// Turning auto_send on, or moving the schedule to another customer,
@@ -19305,7 +19306,7 @@ export const tools: McpTool[] = [
.eq('id', parsedChanges.customer_id)
.eq('company_id', companyId)
.maybeSingle()
if (targetError) throw new Error(`Database error: ${targetError.message}`)
if (targetError) throw dbError(targetError)
if (!target) throw new Error('Customer not found. Use gnubok_list_customers to find IDs.')
if (effectiveAutoSend && !target.email) {
throw new Error('Customer has no email address: auto_send requires one.')
+110
View File
@@ -0,0 +1,110 @@
/**
* The database-error wrapper, and the classification it exists to unlock.
*
* The idiom `throw new Error(\`Database error: ${error.message}\`)` kept the
* prose and dropped `code`. `isTransientFailure()` checks that SQLSTATE FIRST,
* and 57014 (statement timeout) is already in its transient set, so stripping
* it turned a retryable timeout into UNKNOWN_ERROR: "Något gick fel. Försök
* igen." Agents cannot dispatch on that, so they retried: on production over
* 60 days, 462 wasted repeat calls, with 53.1% of all real-agent error calls
* sitting inside a repeat streak.
*
* The headline test is the last one: it asserts the OLD shape still resolves
* to UNKNOWN_ERROR and the new one resolves to TRANSIENT_ERROR, so it fails if
* the wrapper ever stops preserving the code.
*/
import { describe, it, expect } from 'vitest'
import { dbError, errorCauseTag } from '../db-error'
import { getStructuredError } from '../get-structured-error'
/** What supabase-js hands back when Postgres cancels on statement_timeout. */
const TIMEOUT_ERROR = {
message: 'canceling statement due to statement timeout',
code: '57014',
details: null,
hint: null,
}
describe('dbError', () => {
it('preserves the SQLSTATE, which is the whole point', () => {
const wrapped = dbError(TIMEOUT_ERROR)
expect(wrapped.code).toBe('57014')
})
it('preserves details and hint for the server log', () => {
const wrapped = dbError({
message: 'duplicate key value violates unique constraint',
code: '23505',
details: 'Key (company_id, account_number) already exists.',
hint: 'Use upsert.',
})
expect(wrapped.details).toContain('already exists')
expect(wrapped.hint).toBe('Use upsert.')
})
it('never renders the literal "undefined"', () => {
// A driver-level failure (aborted fetch, gateway timeout) can arrive with
// no message at all. "Database error: undefined" is the string that made
// these unsearchable in production.
for (const shape of [{}, { message: undefined }, { message: '' }, { message: ' ' }, null]) {
expect(dbError(shape).message).not.toContain('undefined')
}
})
it('prefixes with the historical context by default', () => {
expect(dbError({ message: 'boom' }).message).toBe('Database error: boom')
})
it('keeps the driver message verbatim when context is null', () => {
// fetchAllRows passes null: callers such as query_journal's
// sanitizeDbError already match on the exact text, and this change is
// meant to add the code, not reword anything.
expect(dbError({ message: 'boom' }, null).message).toBe('boom')
})
it('accepts a custom context', () => {
expect(dbError({ message: 'boom' }, 'Database error resolving voucher "A-7"').message).toBe(
'Database error resolving voucher "A-7": boom',
)
})
})
describe('errorCauseTag', () => {
it('returns the SQLSTATE, which carries no tenant data', () => {
expect(errorCauseTag(dbError(TIMEOUT_ERROR))).toBe('57014')
})
it('falls back to the error name when there is no code', () => {
expect(errorCauseTag(new TypeError('nope'))).toBe('TypeError')
})
it('returns null rather than inventing a tag', () => {
expect(errorCauseTag(new Error('plain'))).toBeNull()
expect(errorCauseTag(null)).toBeNull()
})
})
describe('classification: the regression this prevents', () => {
it('classifies a wrapped statement timeout as retryable', () => {
expect(getStructuredError(dbError(TIMEOUT_ERROR)).code).toBe('TRANSIENT_ERROR')
})
it('would classify the OLD bare-Error shape as UNKNOWN_ERROR', () => {
// The bug, pinned. `new Error(error.message)` is what fetchAllRows threw,
// and PostgREST does not always put the word "timeout" in the message it
// returns, so message-pattern matching cannot be relied on. Only the code
// is durable, which is why the wrapper must carry it.
const stripped = new Error('some driver text that names no timeout')
expect(getStructuredError(stripped).code).toBe('UNKNOWN_ERROR')
// Same failure, code intact: now dispatchable.
const preserved = dbError({ message: 'some driver text that names no timeout', code: '57014' })
expect(getStructuredError(preserved).code).toBe('TRANSIENT_ERROR')
})
it('does not mistake a SQLSTATE for one of our own stable codes', () => {
// extractCode() only accepts /^[A-Z_]+$/. Every SQLSTATE contains digits,
// so attaching one cannot hijack the application error registry.
expect(getStructuredError(dbError({ message: 'x', code: '23505' })).code).not.toBe('23505')
})
})
+97
View File
@@ -0,0 +1,97 @@
/**
* Wrap a Supabase/PostgREST error without throwing away its identity.
*
* ## The bug this exists to prevent
*
* The idiom across the MCP tools was:
*
* if (error) throw new Error(`Database error: ${error.message}`)
*
* which keeps the prose and discards `code`. That matters because `code` is
* the SQLSTATE, and `isTransientFailure()` in lib/errors/get-structured-error.ts
* checks it FIRST: `57014` (statement timeout), `40001`, `40P01`, `53300` and
* friends are already in its TRANSIENT_SQLSTATES set. Strip the code and a
* retryable timeout arrives as an anonymous Error, misses every transient
* check, and resolves to UNKNOWN_ERROR: "Något gick fel. Försök igen."
*
* Measured on production over 60 days (bot actors excluded): 1 024 real-agent
* tool failures, 645 of them UNKNOWN_ERROR across 60 actors and 57 companies,
* 537 carrying that exact generic string. `gnubok_query_journal` alone failed
* 164 times at a p50 of 8 110 ms while every other failing tool sat between 1
* and 315 ms: a timeout signature that should have been TRANSIENT_ERROR all
* along. Agents cannot dispatch on "something went wrong", so they retried:
* 82 streaks of three or more identical failures, 462 wasted repeat calls,
* 53.1% of all real-agent error calls sitting inside a streak.
*
* ## Why attaching `code` is safe
*
* `extractCode()` only treats a code as an application error code when it
* matches /^[A-Z_]+$/. Every SQLSTATE contains digits (`57014`, `42P01`,
* `23505`), and so does every PostgREST code (`PGRST200`), so none of them can
* be mistaken for one of our own stable codes. The only behaviour this unlocks
* is the transient check that was always meant to run.
*/
/** The shape of a PostgrestError, narrowed to what we read. */
interface DatabaseErrorLike {
message?: string | null
code?: string | null
details?: string | null
hint?: string | null
}
/**
* An Error carrying the driver's SQLSTATE and diagnostics.
*
* `details` and `hint` are preserved for the server log, not for the agent:
* `getStructuredError` never reads them, so they cannot leak into a tool
* result. They are what makes a production failure debuggable after the fact.
*/
export interface DatabaseError extends Error {
code?: string
details?: string
hint?: string
}
/**
* @param error the `error` half of a supabase-js `{ data, error }` result
* @param context prefix for the message; defaults to the historical
* "Database error" so existing message-pattern matching in
* `inferCode()` keeps working unchanged. Pass `null` to keep
* the driver's message verbatim, for call sites whose exact
* text callers already depend on.
*/
export function dbError(error: unknown, context: string | null = 'Database error'): DatabaseError {
const source = (error ?? {}) as DatabaseErrorLike
const raw = typeof source.message === 'string' && source.message.trim() ? source.message.trim() : null
// Never render the literal "undefined". A driver-level failure (an aborted
// fetch, a gateway timeout) can arrive with no `message` at all, and
// "Database error: undefined" is the string that made these unsearchable in
// the first place.
const prefix = context ?? ''
const message = raw
? (prefix ? `${prefix}: ${raw}` : raw)
: `${prefix || 'Database error'}: no message from the database driver`
const wrapped = new Error(message) as DatabaseError
if (typeof source.code === 'string' && source.code) wrapped.code = source.code
if (typeof source.details === 'string' && source.details) wrapped.details = source.details
if (typeof source.hint === 'string' && source.hint) wrapped.hint = source.hint
return wrapped
}
/**
* A PII-safe identifier for what failed, for telemetry.
*
* A SQLSTATE is five characters of protocol vocabulary and carries no tenant
* data. A raw driver message can quote row values in a constraint violation,
* so it belongs in the server log, never in `event_log`.
*/
export function errorCauseTag(error: unknown): string | null {
if (error === null || error === undefined) return null
const source = error as DatabaseErrorLike & { name?: unknown }
if (typeof source.code === 'string' && source.code) return source.code
if (error instanceof Error && error.name && error.name !== 'Error') return error.name
return null
}
+25
View File
@@ -41,6 +41,31 @@ describe('fetchAllRows', () => {
).rejects.toThrow('boom')
})
it('propagates the driver SQLSTATE so a timeout stays dispatchable', async () => {
// This is the highest-traffic error strip point in the codebase: every
// paginated read goes through it. It used to throw `new Error(msg)`, which
// dropped `code`, so a statement timeout (57014) reached the structured
// error layer anonymous and resolved to UNKNOWN_ERROR instead of the
// retryable TRANSIENT_ERROR. gnubok_query_journal failed this way 164
// times in 60 days at a p50 of 8 110 ms.
const failure = fetchAllRows<Row>(() =>
Promise.resolve({
data: null,
error: { message: 'canceling statement due to statement timeout', code: '57014' },
}),
)
await expect(failure).rejects.toMatchObject({ code: '57014' })
// And the message stays verbatim: query_journal's sanitizeDbError matches
// on the existing text.
await expect(failure).rejects.toThrow('canceling statement due to statement timeout')
})
it('does not render "undefined" when the driver sends no message', async () => {
await expect(
fetchAllRows<Row>(() => Promise.resolve({ data: null, error: { code: '08006' } as never })),
).rejects.toThrow(/^(?!.*undefined).*$/)
})
it('returns [] when the first page is empty', async () => {
const out = await fetchAllRows<Row>(pagedQuery({ 0: [] }))
expect(out).toEqual([])
+14 -1
View File
@@ -1,4 +1,5 @@
import { createLogger } from '@/lib/logger'
import { dbError } from '@/lib/errors/db-error'
const log = createLogger('fetch-all')
@@ -71,7 +72,19 @@ export async function fetchAllRows<T>(
while (true) {
const { data, error } = await queryFn({ from, to: from + PAGE_SIZE - 1 })
if (error) throw new Error(error.message)
// dbError, not `new Error(error.message)`: the bare form discarded the
// driver's SQLSTATE, and `isTransientFailure()` checks that code FIRST.
// A statement timeout (57014) therefore arrived as an anonymous Error and
// resolved to UNKNOWN_ERROR ("Något gick fel. Försök igen.") instead of
// the retryable TRANSIENT_ERROR. This is the single highest-traffic strip
// point in the codebase: every paginated read goes through it, including
// gnubok_query_journal, which failed 164 times in 60 days at a p50 of
// 8 110 ms while every other failing tool sat between 1 and 315 ms.
//
// The message is kept VERBATIM (context: null). Callers such as
// query_journal's sanitizeDbError already match on it, and this change is
// meant to add the code, not to reword anything.
if (error) throw dbError(error, null)
if (!data || data.length === 0) break
allRows.push(...data)
pages += 1