finance-engines-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| product_marginsB | Compute per-product revenue, cost, and margin per metric tonne. Revenue/MT = sum over metals of (assay% x payable% x index price), marked to market against inventory. Returns one record per product with metal contributions and inventory valuation. |
| price_sensitivityB | Margin-per-tonne sensitivity for one product under price shocks: an 'all metals' row plus one row per metal, showing margin/MT under each configured shock (default -25%/-10%/+10%/+25%). |
| breakevenA | Implied per-metal index prices at which a product's margin/MT hits zero (uniform cost/revenue multiple applied to current prices). |
| evaluate_contractsA | Evaluate all configured offtake/feedstock contract structures at spot. Supports grade_multiplier, discount_profit_share, collar, and assay_payables. Returns each contract's outputs and binding flags (collar floor/ceiling binding, profit-share triggered). |
| parse_trial_balanceA | Flatten a Xero Reports/TrialBalance payload into netted section balances (debit positive for assets/expenses, credit positive for revenue/liabilities/equity). Returns a dict keyed by section, each mapping account name -> balance. |
| compute_covenant_metricsB | Compute financial metrics used by loan covenants from a trial balance: revenue, net_income, ebitda, cash, current_ratio, dscr, debt_to_net_worth, total_debt, equity. |
| evaluate_covenantsB | Test computed metrics against covenant thresholds. Each result reports the metric value, operator, threshold, compliant flag, and percentage headroom to the threshold. |
| compliance_certificateA | Generate a full markdown covenant compliance certificate for a period. End-to-end: computes metrics, evaluates covenants, and renders the signable lender certificate in one call. |
| audit_invoicesA | Run all anomaly rules over a set of invoices (and optional line items): duplicate numbers, entry lag, overdue-unpaid, amount outliers, unit-rate changes, new charge types, unexplained credits, and inconsistent tax. Header-level rules always run; item-level rules run only when line items are supplied. |
| normalize_invoice_numberA | Normalize an invoice number for duplicate detection: strips punctuation, uppercases, and drops common prefixes so 'INV123', '#INV123', and 'NV123' collide to the same canonical form. |
| classify_ap_exceptionsA | Classify vendor invoices into the AP exception taxonomy: duplicate invoice (reuses audit normalize/duplicate rules), missing/mismatched PO, missing receipt, wrong legal entity, tax review, and ownerless approval. Each exception includes confidence and a stable reason code. Invoices without match-context fields are only eligible for duplicate and inconsistent-tax reuse. |
| close_readinessA | Inventory ERP extracts, flag stale inputs, report reconciliation coverage, and compute hours-to-close from freeze (or earliest extract) to sign-off or the supplied |
| five_day_closeB | Run the six-gate five-day close reference: source freeze, subledger cutoffs, reconciliation queue, evidence bundle, exception SLA, and controller sign-off. Classifies AP exceptions when invoices are supplied, optionally flashes covenants, and returns operator metrics (straight-through rate, double-handling minutes, stale-input rate, hours-to-close). |
| uipath_handoffB | Accept a UiPath payload and route it to the matching deterministic engine: invoice audit, covenant certificate, contract evaluation, AP exception classification, or five-day close. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 14 tools
Tools cluster around distinct sub-domains (margins, covenants, invoices, close) with clear delineation in descriptions. Some overlap exists (e.g., product_margins vs price_sensitivity, evaluate_covenants vs compliance_certificate), but each tool's scope is explicit enough to avoid misselection.
All names are snake_case and mostly readable, but verb usage is inconsistent: many start with a verb (compute_covenant_metrics, evaluate_covenants, audit_invoices) while others start with a noun (product_margins, breakeven, compliance_certificate, close_readiness). This mixed convention is understandable but not uniformly predictable.
14 tools sit within the ideal 3-15 range and cover multiple finance engines without feeling bloated. Each tool has a distinct purpose, and the breadth of domains (margins, covenants, invoices, close, contracts) justifies the count, though it edges toward the upper bound for such a server.
The tool surface covers major workflows end-to-end: margin analysis (compute, sensitivity, breakeven), covenant handling (metrics, evaluation, certificate), invoice processing (audit, normalization, AP classification), close execution (readiness, five-day), contract evaluation, and trial balance parsing. Minor gaps exist (e.g., no tool to reference available products or list covenant definitions), but for an analytics engine the coverage is strong.