Loading ToolBench

Reading the current corpus…

mcp-newbuild MCP server · F · ToolBench

← Browse

mcp-newbuild

github.com/atomno-mcp/mcp-newbuild

MCP for Russian new-builds on the client's EISZHS contract. Method spec is closed; no invented API.

<=2025-11-25STDIO6 tools

47/100

Definition Quality45

Protocol Readiness47

Supportability50

Spec posture

Inferred effective spec <=2025-11-25·current baseline 2026-07-28

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

All 6 tools have identical, vague Russian descriptions that provide no actionable guidance for LLM selection. Descriptions like 'Карточка застройщика: ключ ЕИСЖС клиента или честный отказ' (22-24 chars) fall far below the 50-200 char baseline for A+ tools and fail to explain WHAT the tool does, WHEN to use it, or what it returns. Input schemas are present with basic type definitions and descriptions, but parameter descriptions are minimal (e.g., 'ИНН или ОГРН застройщика (юрлица)' lacks format/constraint guidance). No output schemas are documented. Error handling is absent from tool definitions. Tool names follow verb_noun convention (check_*, get_*, list_) but descriptions do not compensate for their brevity. The server relies on external documentation (closed EISZHS contract spec) rather than self-documenting tool interfaces.

Tools (6)

check_construction_permitread onlyauthsource verified43/100

Разрешение на строительство: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто.

check_developerread onlyauthsource verified45/100

Карточка застройщика: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто — запрос не выдумываем.

check_escrowread onlyauthsource verified43/100

Эскроу: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто.

get_developer_risk_summaryread onlyauthsource verified45/100

Сводка риска: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто.

get_project_declarationread onlyauthsource verified45/100

Проектная декларация: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто.

list_new_buildingsread onlyauthsource verified50/100

Каталог объектов: ключ ЕИСЖС клиента или честный отказ. Описание методов закрыто.

Issues (5)

CRITICALtool-description ↗

All tool descriptions are under 25 characters and provide no actionable context. These fail to answer WHAT the tool does, WHEN to use it, or what it returns, forcing LLMs to guess.

check_developerget_project_declarationcheck_escrowcheck_construction_permitlist_new_buildingsget_developer_risk_summary

CRITICALresponse-shaper ↗

No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields without knowing response structure.

check_developerget_project_declarationcheck_escrowcheck_construction_permitlist_new_buildingsget_developer_risk_summary

HIGHconstrained-input ↗

Parameter descriptions lack format guidance, constraints, and examples. E.g., 'ИНН или ОГРН застройщика' does not specify format (10 vs 12 digits), valid ranges, or when to use each.

Recommendations

Expand all tool descriptions to 80-150 characters. Example: 'check_developer' → 'Retrieve developer (builder) profile and risk indicators from the official EISZHS registry. Requires valid INN or OGRIN. Returns company status, escrow compliance, and construction permits. Requires EISZHS client credentials.'
Document output schemas for all tools. Specify which fields are always present, which are conditional, and which IDs/references are needed for chaining (e.g., object_id for follow-up calls).
Add format constraints to parameters: 'ИНН: 10 digits (individual) or 12 digits (entity); ОГРН: 13 digits. Example: 7701234567 (INN) or 1027700000001 (OGRIN).'
Document error conditions and recovery paths in tool descriptions: 'Returns {ready: false, error: source_not_configured} if EISZHS credentials are missing. Caller must set MCP_NEWBUILD_EISZHS_CLIENT_ID and MCP_NEWBUILD_EISZHS_CLIENT_SECRET environment variables.'
Add pagination guidance to list_new_buildings: 'Results are limited to 20 items by default (max 100). Use limit parameter to control page size. No cursor/offset provided; results are deterministic for the same filters.'

Spec posture evidence

Inferred effective spec: <=2025-11-25.

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

Score history

Overall score trend

First recorded score · v2 rubric

47/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	47	<=2025-11-25	v2