Jakob WennbergandClaude Opus 4.8
43007fa869
* feat(mcp): make the agent surface self-describing (staging _meta, company identity, clean summaries)
A pass over the MCP server's agent-facing surface so an agent can act
correctly without parsing description prose:
- Machine-readable staging contract: deriveToolMeta() attaches _meta to
tools/list (and search detail=full) — { requires_approval, approve_tool,
preflight? } — keyed off the STAGED_OPERATION_SCHEMA output schema. Literal
_meta (e.g. UI widget hints) wins on collision. TOOL_PREFLIGHT_MAP names the
read-only pre-flight for the few writes that have one (year-end readiness,
VAT validate, depreciation proposal). Guarded by staging-meta.test.ts.
- Company identity in gnubok_get_agent_briefing: returns a `company` block
(id, name, org_number, entity_type, accounting_method) so the agent can
confirm WHICH entity it operates on and pick the right settlement account
(accrual = credit 1510; cash = debit 19xx) before any write. Best-effort —
a missing row never blocks the briefing. Covered by agent-briefing.test.ts.
- toSummary(): trims the long, keyword-stuffed SKILL.md frontmatter into clean
one-liners for gnubok_list_skills / gnubok_get_agent_briefing so the client
never truncates one mid-sentence; full bodies stay in gnubok_load_skill.
Covered by to-summary.test.ts.
- bank-reconciliation skill: a match/link decision tree (what you have x
whether a verifikat exists) and kontant- vs faktureringsmetoden settlement
accounts.
- Prose/description clarifications: "Stages"/"Stages for approval" on the
link tools; propose_dispositioner/accruals note there is no dedicated MCP
poster; server-info documents _meta and the legacy gnubok_ tool prefix.
All 34 touched MCP tests pass. Merged cleanly on top of #759/#760 (server.ts).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(mcp): make accounting_method description state the full settlement posting
Review (swedish-accounting-compliance): the agent-briefing schema described
accrual as "credit 1510 on payment", which reads as a one-sided entry. Spell
out both sides (payment debits 19xx AND credits 1510) so an agent can't infer a
single-leg posting that violates BFL 5 kap double-entry. Mirrors the precision
already in the bank-reconciliation skill body. Payload-size guard still passes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 (1M context) <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