mapbox-geospatial-operations
Expert guidance on choosing the right geospatial tool based on problem type, accuracy requirements, and performance needs
Security Assessment
About mapbox-geospatial-operations
mapbox-geospatial-operations is an advisory skill that helps an AI assistant choose the right geospatial tool from the Mapbox MCP Server based on what a problem actually requires, rather than defaulting to one approach. It solves a recurring ambiguity in geospatial work: whether a task calls for a pure geometric calculation (fast, offline, Turf.js-based) or a real-world routing/navigation API call (road networks, traffic, travel time).
The skill is built around a decision framework and a set of use-case decision matrices. It distinguishes straight-line ("as the crow flies") distance from road distance and travel time, and maps common user questions to specific tools: distance calculations route to distance_tool, directions_tool, or matrix_tool; proximity and containment route to distance_tool, isochrone_tool, and point_in_polygon_tool; routing and navigation route to directions_tool, optimization_tool, map_matching_tool, and bearing_tool; and area/shape operations route to buffer_tool, isochrone_tool, area_tool, simplify_tool, and centroid_tool. Each matrix explains the reasoning — for example, using routing to build a travel-time isochrone zone and geometric containment to test many addresses against it, or clarifying whether "nearest" means geometric or routed.
It is aimed at developers and AI agents integrating the Mapbox MCP Server who need to pick tools correctly for logistics, delivery-zone, geofencing, and navigation tasks. This is guidance-only, read-only documentation about tool selection; it does not itself execute network calls or handle credentials, and includes an evals file for validating the guidance.
FAQ
What problem does this skill address?
Choosing between offline geometric tools (Turf.js) and Mapbox routing/navigation APIs based on whether a task needs pure spatial math or real-world road-network, traffic, and travel-time data.
How does it decide between geometric and routing tools?
It uses the core question of what the problem requires: straight-line distance, point containment, buffers, centroids, and high-frequency checks go to offline geometric tools; road distance, travel time, traffic-aware routing, and route optimization go to routing APIs.
Can you give a concrete example?
For "are these 200 addresses in our 30-minute delivery zone?" it recommends building the zone with isochrone_tool (routing) and then testing each address with point_in_polygon_tool (geometric), because routing creates travel-time zones and geometry handles fast containment checks.
What are the prerequisites?
Access to the Mapbox MCP Server, which provides the referenced tools such as distance_tool, directions_tool, matrix_tool, isochrone_tool, optimization_tool, and the various geometric operations.
Does the skill run code or make API calls itself?
No. It is decision-support documentation that guides tool selection; the actual execution and any API credentials are handled by the Mapbox MCP Server tools it points to. It also ships an evals file for validating the recommendations.
Install mapbox-geospatial-operations
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
mapbox/mapbox-agent-skills