MCP Server for MLIT (Ministry of Land, Infrastructure, Transport and Tourism) Real Estate Information. Provides access to Japanese real estate transaction data, land prices, urban planning zones, school districts, and related geographic/demographic information.
Server has 9 tools with complete input schemas and descriptions. Naming follows verb_noun pattern (list_, fetch_, summarize_, get_, calculate_). Descriptions are present (avg ~150 chars) and explain WHAT the tool does. However, parameter descriptions lack actionable constraints (ranges, formats, enums). Output schemas are not documented, LLMs cannot predict response structure. Error handling is absent, no recovery guidance or categorization. Tool descriptions mention datasets (XIT001, XPT001) but lack WHEN/WHY context for LLM selection. No tool annotations (readOnlyHint, destructiveHint). Composition is sound, each tool has one responsibility. Security: no secrets in params (good), but no audit logging or permission gates visible.
Fetch land price (地価公示) point data from MLIT dataset XPT002. Supports both GeoJSON and PBF (Protocol Buffer) formats.
Fetch elementary school district (小学校区) tile data from MLIT dataset XKT004. Returns MVT (Mapbox Vector Tile) data encoded as base64.
Fetch real estate transaction points as GeoJSON from MLIT dataset XPT001. Requires XYZ tile coordinates. Large responses (>1MB) are returned as resource URIs.
Fetch aggregated real estate transaction data from MLIT dataset XIT001.
Fetch urban planning zone (都市計画区域) data from MLIT dataset XKT001. Requires z/x/y tile coordinates.
Get internal server statistics including cache hits, misses, and request counts.
Output schemas not documented. LLMs cannot predict response structure (fields, types, nesting). Responses are returned as dict.model_dump(), structure is opaque to the agent.
Parameter descriptions lack actionable constraints. E.g., 'prefecture_code' says '2-digit code' but no enum or pattern. 'year' (1995-2024) lacks min/max. 'zoom level (11-15)' is in description, not schema. LLMs cannot validate inputs without formal constraints.
Inferred effective spec: 2025-06-18+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 54 | 2025-06-18+ | v2 |
Return the list of municipalities within the specified prefecture using MLIT dataset XIT002.
Calculate unit prices (price per square meter and price per tsubo) from real estate transaction data. Returns average, median, and distribution by property type.
Summarize real estate transaction data from MLIT dataset XIT001. Returns aggregated statistics (count, average, median, distribution) instead of raw data. Useful for large datasets.
No error handling or recovery guidance. Tools return dict responses but no error categorization (retryable vs fatal). No guidance on what to do if a tile coordinate is invalid or a year is out of range.
Tool descriptions lack WHEN/WHY context. E.g., 'fetch_transaction_points' mentions XPT001 dataset but does not explain when to use it vs fetch_transactions. LLMs cannot disambiguate similar tools without explicit guidance.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools are read-only but this is not declared. Agents cannot infer safety properties without explicit hints.