← Science & Tech
Active
pythonAILLMRAGmountaineeringactive

Mountaineering Preparation Assistant

A conversational AI assistant for technical mountaineering preparation — route research, conditions, weather, and avalanche forecasts, grounded in real data sources.


Technical mountaineering involves a lot of online prep: finding good routes, checking hut availability, reading conditions reports, looking up the weather (both the forecast and the last few days), cross-referencing topos from multiple sources.

This app is a conversational assistant built on Claude (via the Anthropic API), empowered by a set of tools that give it structured access to the information that actually matters.

View on GitHub →

Why an LLM isn’t enough on its own

Without tools, an LLM simply doesn’t have access to the right information. We solve this with a database of scraped data and access to various services via APIs (meteofrance, camptocamp, etc.). Even with tool access, the LLM needs to be constrained in what judgments it makes. The failure modes of LLMs on mountaineering questions are predictable:

Tools

Stack

Status

Active development. The core loop — conversational route research with live weather and avalanche data — is working. Ongoing and planned work includes: