Loading ToolBench
MCP server for month-end close reconciliation — Bank matching, GL reconciliation, Intercompany checking, Break classification, and Close tracking.
Server has 5 tools with complete input schemas and descriptions. Naming follows verb_noun pattern (match_, reconcile_, check_, classify_, track_). Descriptions are substantive (100-200 chars), explaining WHAT and WHEN. However, parameter descriptions lack detail on constraints, formats, and valid ranges. Output schemas are not documented, LLMs cannot plan downstream calls or extract structured data. No error handling guidance. No tool annotations (readOnlyHint, idempotentHint). CSV input parameters accept free-form strings with no validation hints. Missing per-parameter type constraints (enums, patterns, min/max). Overall: solid foundation but incomplete for production agent use.
Compare intercompany ledger entries between two entities for a shared account. Identifies balance mismatches, flags currency conversion differences separately, and returns a structured mismatch report.
Classify a list of unreconciled items by likely cause: Timing Difference, Missing Entry, Duplicate Posting, Currency Rounding, Intercompany Mismatch, or Unknown. Each item receives a recommended next action.
Match bank statement transactions against a GL cash account ledger using amount equality and date proximity (±3 days). Returns matched pairs, unmatched bank items, unmatched GL items, and a full reconciliation summary with closing balance.
Compare two GL account extracts (prior vs current period, or two sub-ledgers). Identifies new entries, reversed entries, and items present in one extract but not the other. Returns a structured reconciliation statement.
Track month-end close progress across all reconciliation tasks. Returns completion %, overdue items, not-started items, and a HIGH/MEDIUM/LOW risk flag based on deadline proximity and outstanding work.
No output schemas documented. LLMs cannot infer response structure, plan chaining, or extract required fields for downstream calls.
CSV input parameters lack format/constraint documentation. No guidance on required columns, encoding, date format, or delimiter. LLMs will guess and pass malformed CSVs.
No error handling guidance. Tools do not document what errors are retryable, what user action is needed, or how to recover from malformed input.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 58 | 2026-07-28+ | v2 |
No tool annotations (readOnlyHint, idempotentHint). All tools are read-only and idempotent, but LLMs cannot infer this without explicit hints.
Parameter descriptions are minimal. 'Bank statement CSV' and 'GL cash ledger CSV' lack detail on required columns, date format (YYYY-MM-DD?), amount format (decimal places?), or encoding.