Files
accounted/extensions/general/mcp-server
Jakob Wennberg 89d0e1b994 fix(mcp): promote gnubok_update_customer to the default tools/list catalog (#1986)
* fix(mcp): promote gnubok_update_customer to the default tools/list catalog

The tool existed since #1876 but was catalogVisibility 'search', which
means it appeared only through gnubok_search_tools. Claude.ai (and other
list-driven clients) can only call tools present in tools/list, so on the
primary connector the tool was uncallable and the reporter of #1706 read
it as missing twice.

- server.ts: drop the 'search' flag on gnubok_update_customer; no change
  to schema, executor, scopes, risk tier, or staging path
- update-customer.test.ts: pin the tools/list projection
  (isDefaultCatalogTool) instead of the flag; keep the search test
- payload-size.bench.test.ts: ceiling 63_400 to 64_100 (measured 64 043,
  +761 over main) with a progression bullet
- README.md: note that search-only tools are not callable from Claude.ai
- DECISIONS.md: supersede the 2026-08-25 keep-search-only entry

Closes #1706

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkUfWtuFCUkNtRAgMQCse2

* test(mcp): re-measure the tools/list ceiling after #1993 landed on main

#1993 declared line_type and revenue_account on the create item schema,
which moved the accounted projection to 63 761 with gnubok_update_customer
promoted; ceiling 63_800 keeps the same headroom as before.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FkUfWtuFCUkNtRAgMQCse2

---------

Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-08-27 22:28:14 +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 (the SDK is called from lib/ai/provider.ts and lib/ai/services/anthropic-family.ts; features such as extensions/general/invoice-inbox/lib/extract-invoice-fields.ts go through getAiService() in lib/ai rather than the SDK directly): 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 on the wire without bloating the default catalog. Claude.ai only calls tools present in tools/list, so a tool that a user or a skill must call directly stays in 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, registered in resources/index.ts: company/current, period/active, recent-activity, capabilities, attention, chart-of-accounts, settings/vat-treatments, booking-templates, ledger/context, reconciliation/summary
  • widgets/: inline HTML widgets (receipt-matcher, vat-review, pending-operations)
  • prompts/: slash-command-style prompts
  • skills/: domain-knowledge skill bodies served via gnubok_load_skill
  • public-tools.ts: lazy authentication (issue #1814). ANONYMOUS_METHODS (initialize, ping, tools/prompts/resources listing) and the three PUBLIC_TOOLS (gnubok_search_tools, gnubok_list_skills, gnubok_load_skill) answer without credentials, rate-limited per truncated IP; every other tools/call gets a transport-level 401 + WWW-Authenticate from handleMcpRequest in server.ts, which is the challenge clients turn into their Connect prompt
  • tasks.ts: MCP Tasks extension (io.modelcontextprotocol/tasks): durable handles for long-running tool calls, rows in mcp_tasks (service-role writes only)
  • origin-guard.ts: Origin-header validation on the Streamable HTTP endpoint (DNS-rebinding defence required by the MCP spec)
  • staging-pii-guard.ts: refuses a plaintext personnummer in staged pending_operations params/preview, so every staging tool inherits the encrypt-at-staging rule
  • tool-namespace.ts: ?tool_namespace=accounted handling (accounted_* aliases for the canonical gnubok_* ids)
  • __tests__/: strictness guards + per-tool coverage