Loading ToolBench

Reading the current corpus…

1c-odata-mcp MCP server · F · ToolBench

1c-odata-mcp

MCP server for 1C:Enterprise (1С) over OData. Ask your accounting database in natural language — debtors, sales, taxes, cash flow — from any MCP client (Claude, Cursor, VS Code, local models…). 55 tools, run with npx, nothing installed inside 1C; read-only by default with opt-in (gated) write.

2025-06-18+STDIO17 tools

48/100

Definition Quality50

Protocol Readiness47

Supportability45

Spec posture

Inferred effective spec 2025-06-18+·current baseline 2026-07-28

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

17 tools with consistent naming (verb_noun pattern: read.*, write.*) and descriptions present. However, critical gaps: (1) Output schemas are NOT documented anywhere in the codebase, no return type specifications visible for any tool. (2) Parameter descriptions exist but are minimal (avg ~40 chars, well below the 72-char baseline). (3) No error handling guidance, tools return raw API errors without recovery hints. (4) Two destructive tools (post_document, mark_for_deletion) lack explicit confirmation workflows in visible code. (5) Tool annotations are applied generically via a wrapper (READ_HINTS, WRITE_HINTS, DESTRUCTIVE_HINTS) but the actual hint values are not shown in source. This is a mid-tier server with solid naming but missing output documentation and error recovery patterns.

Tools (17)

read.cashflow.summaryread onlyauth50/100

Get cash flow summary by period and category

read.counterparties.getread onlyauth50/100

Get detailed information about a specific counterparty

read.counterparties.listread onlyauth50/100

List counterparties (контрагенты) with basic info

read.counterparties.searchread onlyauth50/100

Search counterparties by name or INN

read.documents.getread onlyauth50/100

Get detailed information about a specific document

read.documents.listread onlyauth50/100

List documents of a specific type

read.organization.bank_accountsread onlyauth50/100

Get bank accounts for an organization

read.organization.inforead onlyauth50/100

Issues (5)

CRITICALpattern:tool ↗

No output schemas documented. Tools return data but LLMs cannot see what fields to expect. This violates pattern:tool and forces agents to guess at response structure.

read.system.list_databasesread.counterparties.listread.documents.listread.registers.stock_balanceread.sales.top_customers

HIGHpattern:tool-description ↗

Parameter descriptions are minimal (avg ~40 chars vs 72-char baseline). E.g., 'Database name' lacks context on format, required vs optional, or how to discover valid values.

read.system.list_organizationsread.counterparties.listread.documents.listread.registers.stock_balance

HIGHpattern:recovery-guide ↗

No error handling guidance in tool descriptions. Tools do not explain what to do if a database/organization/counterparty is not found, or how to recover.

read.counterparties.getread.documents.getwrite.document.post_documentwrite.entity.mark_for_deletion

Recommendations

Document output schemas for all tools. For list tools, specify: returns array of {id, name, ...}, total_count, next_cursor. For get tools, specify all returned fields and their types.
Expand parameter descriptions to 60 - 120 chars. Include format hints: 'Database name (string, case-sensitive; see read.system.list_databases for valid names)' instead of just 'Database name'.
Add error recovery guidance to tool descriptions. E.g., 'If counterparty not found, call read.counterparties.search() with a partial name to discover the correct spelling.'
Make the confirm-then-execute pattern explicit in tool descriptions: 'Set confirm=false (default) to preview changes; set confirm=true to apply. Always show preview first.'
Expose the actual hint values (READ_HINTS, WRITE_HINTS, DESTRUCTIVE_HINTS) in source or document them in a constants file so they can be verified against the spec.

Spec posture evidence

Inferred effective spec: 2025-06-18+.

Relies on Logging (deprecated) - log to stderr or use OpenTelemetry

Score history

Overall score trend

First recorded score · v2 rubric

48/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	48	2025-06-18+	v2