Files
accounted/extensions/general/mcp-server
MattssonandClaude Fable 5 77cacdcf34 feat(mcp): personal_number on gnubok_update_customer (#1876) (#1890)
gnubok_create_customer takes a personnummer (encrypted before approval)
but gnubok_update_customer did not, so an existing customer whose
personnummer sat in the org-number field could not be corrected via MCP.
The REST PATCH already supports it; this closes the MCP/pending-operations
gap across its three layers:

- tool inputSchema: personal_number (string or null) on the strict
  whitelist. The tool validates the plaintext before any DB read and
  mirrors the REST PATCH semantics: masked echo (********-1234 or
  ********-????) = leave unchanged, explicit null = clear, absent =
  untouched. Setting is refused unless the row ends up as an individual
  (GDPR art. 5.1 c), including via a simultaneous type change.
- CustomerChangesSchema: personal_number_encrypted (nullable, ciphertext
  shape per customers_personal_number_check 20260726110000). The
  plaintext key stays forbidden by .strict() and staging-pii-guard.
- update executor: maps the staged ciphertext onto customers
  .personal_number (set/clear/leave), re-checks the individual-only rule
  against a tampered row, and returns only personal_number_masked.

PII handling: the personnummer is encrypted at staging time
(AES-256-GCM, same path as create); pending_operations params carry only
the ciphertext and the approval preview only the masked form. Idempotency
hashing switches to the masked preview for personnummer-bearing updates
(random-IV ciphertext would break retries); other updates keep their
previous hash identity.

catalogVisibility stays 'search': tools/list is at its 59.95K token
ceiling with zero headroom (see DECISIONS.md).

Fixes #1876

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-08-25 14:35:02 +02:00
..
2026-07-24 15:03:50 +02:00
2026-07-24 15:03:50 +02:00
2026-07-24 15:03:50 +02:00

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. Add ?tool_namespace=accounted for the Accounted tool names. Requests without it retain the legacy Gnubok namespace. OAuth and stdio bridges live alongside the API surface: see app/api/mcp-oauth/, packages/accounted-mcp/, and the compatibility package in packages/gnubok-mcp/.

Tool authoring contract

Enforced by tests in __tests__/: these are not style preferences, they're guard rails.

  1. additionalProperties: false on every inputSchema. Guarded by strict-schemas.test.ts. Forces clear rejections on hallucinated fields instead of silent ignores.
  2. Descriptions ≤ 280 chars. Guarded by output-schema.test.ts. No Args: / Returns: / Examples: prose: those belong in JSON Schema. Use agent-native hints ("Use to…", "Call X first", "HIGH risk").
  3. Staged-operation envelope for write tools: outputSchema: STAGED_OPERATION_SCHEMA (server.ts). Fields: staged, risk_level, actor, message, preview, period_status?, next?. The staged: true boolean is the explicit completion signal; agents must not infer completion from prose. Do NOT introduce a parallel { success, shouldContinue, output } envelope.
  4. period_status threading: any tool that ties to a fiscal-period-bound date (categorize, mark paid, create voucher, correct/reverse entry, approve supplier invoice) passes dateForPeriodCheck to stagePendingOperation. Response then includes period_status: { period_id, status: open|locked|closed, lock_date } so widgets and agents disable writes without round-trips.
  5. Scope mapping: every new tool needs an entry in lib/auth/api-keys.ts TOOL_SCOPE_MAP. Missing entries default to deny.
  6. Tests for new write tools: add staging-gate coverage to __tests__/voucher-tools.test.ts (or a sibling) plus executor coverage to lib/pending-operations/__tests__/voucher-executors.test.ts if the tool stages a new operation_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. If the test fires, the right answer is rarely "raise the ceiling". Instead, trim descriptions or set specialized wide tools to catalogVisibility: 'search'. Those tools remain discoverable with full schemas through gnubok_search_tools and callable through tools/call without bloating the default catalog.

Where things live

  • server.ts: the tools array + JSON-RPC dispatcher
  • tool-result.ts: withNext(), toToolError() response helpers
  • resources/: read-only Accounted:// 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 prompts
  • skills/: domain-knowledge skill bodies served via gnubok_load_skill
  • __tests__/: strictness guards + per-tool coverage