Jakob WennbergandClaude Fable 5
21512db81a
* feat(mcp): unify missing-document surfaces on one predicate (P1-3) The two MCP surfaces told different truths: the transactions tool keyed 'has underlag' on transactions.document_id while the verifikat tool keyed on document_attachments — and neither respected the source-type semantics, version chains, or journal_entry_no_doc_required waivers that lib/worklist's canonical count applies. Measured on prod: 22,046 waived verifikat still listed to agents, 2,370 doc-exempt source types listed, ~87 docs attached to transactions but never propagated to the verifikat, 1,100 transactions flagged missing-receipt although their verifikat HAS the underlag. One predicate now lives in SQL — posted, needs-doc source type (mirrors NEEDS_DOC_SOURCE_TYPES), no current-version doc, no waiver: - verifikat_without_documents RPC v2 adopts the canonical predicate. - New transactions_without_documents RPC: the bank-driven subset of the same predicate, joined through transactions.journal_entry_id — a strict subset of the verifikat surface by construction. Rows expose qualified transaction_id (P1-2 forward-compat); bare id deprecated. - Both tools become thin RPC wrappers; descriptions state the actual set relationship. - lib/worklist countVerifikatMissingDocument delegates to the RPC (previously three full-table pulls set-differenced client-side) — badge count and agent surfaces can no longer drift. - Backfill: propagate transaction-attached docs to their verifikat where the attachment was never linked (open periods only; never steals a doc linked to another verifikat). pg-real: fixture matrix (no-doc/with-doc/waived/stale-version/ doc-exempt-source/import), strict-subset assertion, per-source-type pin of the SQL list against the TS constant, tenant guard. Part of dev_docs/mcp_optimization_plan.md (P1-3). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs(mcp): explicit grants restated + count-call comment (#876 review) - Restate REVOKE/GRANT on verifikat_without_documents so the migration is self-contained (CREATE OR REPLACE preserves the 20260703130000 grants — verified on prod: authenticated + service_role only). - Comment on the p_limit:1 count call: total_count is computed over the full filtered set, independent of page size. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Accounted MCP server
JSON-RPC 2.0 server exposing the Accounted bookkeeping engine to MCP clients (Claude Desktop, Claude Code, etc.). Endpoint: /api/extensions/ext/mcp-server/mcp. OAuth and stdio bridge live alongside the API surface — see app/api/mcp-oauth/ and packages/gnubok-mcp/.
Tool authoring contract
Enforced by tests in __tests__/ — these are not style preferences, they're guard rails.
additionalProperties: falseon everyinputSchema. Guarded bystrict-schemas.test.ts. Forces clear rejections on hallucinated fields instead of silent ignores.- Descriptions ≤ 280 chars. Guarded by
output-schema.test.ts. NoArgs:/Returns:/Examples:prose — those belong in JSON Schema. Use agent-native hints ("Use to…", "Call X first", "HIGH risk"). - Staged-operation envelope for write tools —
outputSchema: STAGED_OPERATION_SCHEMA(server.ts). Fields:staged, risk_level, actor, message, preview, period_status?, next?. Thestaged: trueboolean is the explicit completion signal; agents must not infer completion from prose. Do NOT introduce a parallel{ success, shouldContinue, output }envelope. period_statusthreading — any tool that ties to a fiscal-period-bound date (categorize, mark paid, create voucher, correct/reverse entry, approve supplier invoice) passesdateForPeriodChecktostagePendingOperation. Response then includesperiod_status: { period_id, status: open|locked|closed, lock_date }so widgets and agents disable writes without round-trips.- Scope mapping — every new tool needs an entry in
lib/auth/api-keys.tsTOOL_SCOPE_MAP. Missing entries default to deny. - Tests for new write tools — add staging-gate coverage to
__tests__/voucher-tools.test.ts(or a sibling) plus executor coverage tolib/pending-operations/__tests__/voucher-executors.test.tsif the tool stages a newoperation_type.
Determinism / cache stability
Tool definitions (name, description, inputSchema, outputSchema, annotations) are declared as static object literals at module load — no timestamps, no UUIDs, no Date/Math.random in the definition layer. This makes the tools/list JSON payload byte-stable across requests, which lets agent-side prompt caches stay warm. Do not introduce per-request non-determinism into the definitions block. Anything time-bound or random belongs inside execute().
For internal Anthropic API usage (today only extensions/general/invoice-inbox/lib/extract-invoice-fields.ts): annotate stable prefixes with cache_control: { type: 'ephemeral' } and log usage.cache_read_input_tokens for hit-ratio observability. The 1h TTL from the agent-native API plan (item 10) requires the direct Anthropic API; Accounted's Bedrock path defaults to a shorter TTL.
Payload-size watchdog
payload-size.bench.test.ts enforces a tools/list JSON payload ceiling (currently 36,000 tokens — bumped from 32,000 when top-level Tool.title landed on all tools for Claude Connectors Directory readiness). If the test fires, the right answer is rarely "raise the ceiling" — instead, trim descriptions or leverage gnubok_search_tools (already deployed; tool definitions can defer to it for discovery rather than enumerating in tools/list).
Where things live
server.ts— the tools array + JSON-RPC dispatchertool-result.ts—withNext(),toToolError()response helpersresources/— read-onlyAccounted://URIs (active company, period, recent activity, capabilities, attention items, voucher gaps, chart of accounts, VAT treatments)widgets/— inline HTML widgets (receipt-matcher, vat-review)prompts/— slash-command-style promptsskills/— domain-knowledge skill bodies served viagnubok_load_skill__tests__/— strictness guards + per-tool coverage