Files
accounted/extensions/general/mcp-server
Jakob WennbergandClaude Fable 5 205610d200 feat(mcp): distribution-channel client marker in MCP telemetry (#706)
* feat(mcp): distribution-channel client marker in MCP telemetry

Record an optional client marker on mcp.tool_called, mcp.tools_list_called
and mcp.resource_read events so per-channel adoption (e.g. the OpenClaw
skill) is measurable in event_log (180-day TTL).

- Server reads X-Gnubok-Client header, falling back to a ?client= query
  param on the endpoint URL. Sanitized ([A-Za-z0-9._-]{1,64}, lowercased),
  telemetry-only — same trust level as Mcp-Session-Id, never auth.
- The query param works with the already-published gnubok-mcp 1.0.1 via
  GNUBOK_URL, so no npm release is required to start measuring.
- Bridge 1.1.0 additionally forwards GNUBOK_CLIENT as X-Gnubok-Client.

OAuth-path attribution via DCR client_name is a possible follow-up — DCR
is stateless today, so client_name isn't recoverable at token time.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(mcp): address PR #706 compliance findings

- ropa.yaml: declare the distribution-channel marker in the mcp.telemetry
  processing activity (GDPR Art. 30 — RoPA was drifting from actual flow)
- bridge: mirror the server's allow-list on GNUBOK_CLIENT so an invalid
  value degrades to no header instead of fetch() rejecting every request
- lib/events/types.ts: annotate client as client-supplied/telemetry-only
- test: pin that the allow-list runs on the percent-decoded ?client= value

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-06-10 14:33:24 +02:00
..
2026-05-04 11:12:29 +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. 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.

  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 (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 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