Loading ToolBench

Reading the current corpus…

GPT-MAPS MCP server · D · ToolBench

GPT-MAPS

An MCP server that provides tools to build interactive map scenes with markers, paths, polygons, and labels using geographic data from OpenStreetMap and routing services.

2026-07-28+STREAMABLE_HTTP5 tools

59/100

Definition Quality55

Protocol Readiness68

SupportabilityNot measured-

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence · evidence: Streamable HTTP

Current-spec patterns detected

Summary

Server has 5 tools with basic schemas and descriptions, but significant gaps limit production readiness. Tool names follow verb_noun convention (createLabel, createMarker, createPath, createPolygon, renderScene), which is good. However, descriptions are minimal (10-50 chars), parameter descriptions lack detail about constraints and formats, and output schemas are not documented. No error handling guidance, no pagination support, and no tool annotations. The monorepo structure with shared services is well-organized, but tool definitions themselves lack the depth needed for reliable LLM invocation.

Tools (5)

createLabelwritesource verified58/100

Create a text label on the map at a location.

createMarkerwritesource verified55/100

Create a marker on the map.

createPathwrite50/100

Create a path or route between two locations on the map.

createPolygonwrite50/100

Create a polygon to highlight a region or boundary on the map.

renderSceneread onlysource verified47/100

Render the complete map scene with all created markers, paths, polygons, and labels.

Issues (5)

HIGHtool-description ↗

Tool descriptions are too brief (10-50 chars). LLMs cannot determine when to select each tool or what distinguishes createLabel from createMarker. Descriptions should be 50-200 chars explaining WHAT, WHEN, and WHAT IT RETURNS.

createLabelcreateMarkercreatePathcreatePolygonrenderScene

HIGHconstrained-input ↗

Parameter descriptions lack constraint details. 'location' is described as 'geographic location' but does not specify format (lat/lon, address string, place name). 'style' object properties have minimal guidance on valid color formats or numeric ranges.

createLabelcreateMarkercreatePathcreatePolygon

HIGHresponse-shaper ↗

Output schemas are not documented. Tools return objects but LLMs cannot see what fields are in the response, what IDs are returned, or how to chain results to downstream tools. renderScene output is completely undocumented.

createLabelcreateMarkercreatePath

Recommendations

Expand tool descriptions to 80-150 chars. Example: 'createLabel: Add a text label to the map at a specific location. Useful for annotating points of interest. Returns label_id for reference.' This tells LLMs WHAT, WHEN, and WHAT IT RETURNS.
Document parameter constraints in descriptions. For 'location', specify: 'Geographic location as address string (e.g., "New York, NY"), place name, or lat/lon coordinates (e.g., "40.7128,-74.0060").' For numeric style properties, add ranges: 'fontSize: number between 8 and 72 pixels.'
Add output schema documentation to each tool. Example for createLabel: 'Returns {label_id: string, location: {lat: number, lon: number}, text: string}'. This enables LLMs to chain results and extract needed IDs.
Define error scenarios and recovery guidance. Example: 'If location is invalid, returns error with message "Location not found. Try a different address or coordinates." Suggest search_locations() tool if available.'
Add tool annotations (destructiveHint, idempotentHint) to indicate which tools modify state. createLabel, createMarker, createPath, createPolygon are WRITE operations; renderScene is READ_ONLY. Annotations help agents plan safely.

Score history

Overall score trend

First recorded score · v2 rubric

59/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	D	59	2026-07-28+	v2