Mountaya

Developers

Get started
AI tools

Agent skills

Agent skills

Drop-in agent skills that teach AI coding assistants how to use Mountaya: authentication, the Data API, the Tile API, and Map Embedding.

Mountaya publishes a set of agent skills in the mountayaapp/agent-skills repository. Each skill is a focused package of instructions (and, for some, helper scripts) that an AI assistant loads on demand. Install them once and your agent knows how to authenticate, compute routes, render tiles, and embed the studio without you pasting docs into the chat.

If you want an agent to call the Data API directly over a live protocol instead of generating client code, see the MCP server, which exposes the same Data API as read-only MCP tools.

Skills

mountaya-auth

API authentication: publishable keys, session tokens, and rate limits. This is the prerequisite skill; the Data API skill depends on it for session-token creation.

mountaya-data-api

The Data API for outdoor routing and geospatial analysis: directions, suggestions, isochrones, matrices, geocoding, POIs, search, reading a saved itinerary and exporting it as a route file, geometry analysis, and terrain — elevation, slope, and aspect at a point, along a route, or across an area. Requires mountaya-auth.

mountaya-tiles-api

The Tile API: rendering overlays, itineraries, and collections with MapLibre GL JS.

mountaya-embedding

Map Embedding: iframe integration, parameters, postMessage events, and safety rules.

mountaya-mcp

The MCP server: connecting Claude, ChatGPT, Cursor, or VS Code over OAuth or an API key pair, the read-only tools, and connector troubleshooting.

Skills are instructions you install into a coding agent. If your assistant connects over MCP instead, the server offers the equivalent procedures as prompts — no install step, listed by the client alongside the tools.

Install

The skills install from the public repository. The CLI is the recommended path; manual copying works for any tool that reads a skills directory.

Install the whole set, or pass --skill to install one at a time. mountaya-data-api depends on mountaya-auth, so install both when you use the Data API skill.

npx skills add mountayaapp/agent-skills

Install one at a time with --skill:

npx skills add mountayaapp/agent-skills --skill mountaya-auth
npx skills add mountayaapp/agent-skills --skill mountaya-data-api
npx skills add mountayaapp/agent-skills --skill mountaya-tiles-api
npx skills add mountayaapp/agent-skills --skill mountaya-embedding
npx skills add mountayaapp/agent-skills --skill mountaya-mcp

Setup

The script-backed skills read your Mountaya API keys from the environment. Set them once so agents can authenticate.

export MOUNTAYA_PUBLISHABLE_KEY="pk_your_publishable_key"
export MOUNTAYA_SECRET_KEY="sk_your_secret_key"
MOUNTAYA_PUBLISHABLE_KEY

pk_…

Your publishable key. Identifies your organization and is required for all products. Enable the scopes the skills need: data for the Data API, tiles for the Tile API, resources for reading and writing itineraries, collections and zones, and embedding for Map Embedding (which enables tiles and resources with it). The embedding scope also gates the studio co-pilot's own gateway, so a key without it cannot start an assistant run.

MOUNTAYA_SECRET_KEY

sk_…

Your secret key. Used to create session tokens for the Data API. Keep it server-side only.

Create keys in your Mountaya app, by opening your organization and its API keys tab, and enable the scopes each skill needs. The skills read these variables when set, and otherwise fall back to a pk_… placeholder you replace by hand.

See also

MCP server

Call the Data API directly from an AI agent over the Model Context Protocol, with read-only tools.

Authentication

How the publishable and secret key pair works, scopes, session tokens, and rate limits.

Previous

MCP server