Mountaineering Skills for Claude Code

Automated mountain route research for North American peaks, built for Claude Code.

Quick Start • How It Works • Features • Installation • Support

Ask Claude to research any mountain. The route-researcher skill pulls from 10+ mountaineering sources and compiles a detailed Markdown report with current weather, avalanche conditions, daylight windows, trip reports, and route beta. What used to take 3-5 hours of tab-hopping now takes 3-5 minutes.

Other projects you might like: PNW Climb Planner · mountaineers-mcp · mountaineers-assistant · peakbagger-cli

See it in action:

Quick Start

Install with npx skills:

npx skills add dreamiurg/claude-mountaineering-skills

Then just ask naturally: "Research Mount Rainier". Claude generates a route beta report in your current directory.

npx skills installs the route-researcher skill, which drives the full workflow from natural-language requests. If you also want the /mountaineering:* slash commands, install the plugin instead -- see Installation.

How It Works

The skill uses a hybrid architecture: Python scripts for deterministic API calls, LLM agents for tasks requiring judgment.

graph TB
    Start([User asks Claude to<br/>research a peak]) --> Search[Phase 1-2: Peak Identification<br/>Search PeakBagger, fetch details]

    Search --> Parallel[Phase 3: Parallel Data Gathering]

    Parallel --> Python[Python Script<br/>Weather, daylight,<br/>avalanche, air quality]
    Parallel --> Agent1[Researcher Agent 1<br/>PeakBagger + SummitPost]
    Parallel --> Agent2[Researcher Agent 2<br/>WTA + Mountaineers]
    Parallel --> Agent3[Researcher Agent 3<br/>AllTrails]

    Python --> Analyze
    Agent1 --> Analyze
    Agent2 --> Analyze
    Agent3 --> Analyze

    Analyze[Phase 4: Route Analysis<br/>Synthesize data, identify hazards]

    Analyze --> Writer[Phase 5: Report Writer Agent<br/>Generate markdown report]

    Writer --> Reviewer[Phase 6: Report Reviewer Agent<br/>Validate accuracy, fix issues]

    Reviewer --> End([Phase 7: User receives<br/>route beta report])

    style Start fill:#e1f5ff
    style End fill:#e1f5ff
    style Parallel fill:#fff4e1
    style Python fill:#e8f5e9
    style Agent1 fill:#f0f0f0
    style Agent2 fill:#f0f0f0
    style Agent3 fill:#f0f0f0
    style Writer fill:#fff3e0
    style Reviewer fill:#fff3e0

Three researcher agents gather data in parallel while a Python script fetches conditions. Dedicated agents write and review the final report. If a source fails, the skill documents the gap and continues.

Features

Data Sources

The skill aggregates from specialized mountaineering sites:

Coverage note: Report quality depends on how well-documented your peak is across these sources. Works best for popular North American peaks.

Graceful Degradation

Missing data? The skill notes what's unavailable in an "Information Gaps" section and provides manual lookup links. You always get a report, even if some sources are down.

Installation

Prerequisites: Claude Code, Node.js (for npx), optionally uv for Python tools.

Recommended: npx skills

npx skills installs the route-researcher skill directly:

npx skills add dreamiurg/claude-mountaineering-skills

Useful flags:

npx skills add dreamiurg/claude-mountaineering-skills --list   # preview before installing
npx skills add dreamiurg/claude-mountaineering-skills -g        # install globally (all projects)
npx skills add dreamiurg/claude-mountaineering-skills -y        # skip confirmation prompts

This gives you the full natural-language workflow ("Research Mount Rainier"). The /mountaineering:* slash commands are not included -- install the plugin below if you want them.

Alternative: Claude Code plugin

The plugin bundles the same skill plus the /mountaineering:* slash commands:

/plugin marketplace add dreamiurg/claude-mountaineering-skills
/plugin install mountaineering@mountaineering-marketplace
/reload-plugins

Python dependencies install automatically if uv is available. No restart needed -- /reload-plugins activates the plugin in the current session.

Usage

Commands

Natural Language

You can also just ask naturally:

"Research Mt Baker"
"Get route beta for Forbidden Peak"
"I'm planning to climb Sahale Peak, can you research it?"

Reports save to your current directory as YYYY-MM-DD-peak-name.md.

Migration from v4.x

If you previously installed the plugin as mountaineering-skills, reinstall with the new name:

/plugin uninstall mountaineering-skills
/plugin install mountaineering@mountaineering-marketplace

Dependencies

• peakbagger-cli >= v1.10.0 - peak data and trip reports
• Python tools - weather, avalanche, and daylight calculations

Updates

If you installed with npx skills:

npx skills list                       # see installed skills
npx skills update route-researcher    # update to latest

If you installed the plugin:

/plugin list                          # check current version
/plugin update mountaineering         # update to latest

Contributing

Pull requests welcome. See CONTRIBUTING.md.

Support

Open an issue or start a discussion.

License

MIT

More from @dreamiurg

• 🏔️ PNW Climb Planner — pick a Washington peak, see the odds of a climbable day from 20 years of weather data, and line up backups (the story behind it)
• mountaineers-mcp — mountaineers.org for AI assistants: activities, courses, routes, trip reports
• mountaineers-assistant — Chrome extension with your Mountaineers climbing stats, local-only
• peakbagger-cli — search and analyze PeakBagger.com peak data in your terminal
• more at dreamiurg.net/projects

Made by @dreamiurg in Seattle. If this project saved you time, you can buy me a coffee — appreciated, never expected.