CountyEthics
About

About CountyEthics

AI-powered condition analysis for UK climbing crags — protecting fragile rock and keeping climbers safe.

How it works

Every crag is automatically analysed twice daily. Here's what goes into each assessment.

1

Fetch weather data

28 days of history + 7-day forecast from Open-Meteo (UK Met Office model), pinned to the crag's exact coordinates.

More detail
We calculate consecutive dry days, precipitation totals (7-day and 28-day), average humidity and temperature, and forecast rainfall. The last 24 hours of hourly readings (precipitation, temperature, humidity, wind, cloud cover) give the AI granular recent conditions. All of this is shown on each crag's weather tab.
2

Load site characteristics

Rock type, wind exposure, altitude, aspect, and climbing styles — the physical attributes that determine how quickly a crag dries.

More detail
The AI receives a dynamically built system prompt tailored to the crag's rock type, including detailed geology (porosity, water absorption rates, structural failure mechanisms) and site-specific drying context derived from its aspect, wind exposure, and altitude. Porous sandstone gets far more cautious thresholds than hard whinstone or gritstone.
3

Gather climber reports

Recent condition reports from visiting climbers (dry / damp / wet / seeping…) are included. A first-hand observation carries significant weight.

More detail
Reports can include photos, what3words locations, and optional contact details. When weather data alone is ambiguous, a recent report from someone who was actually there can tip the verdict. Anyone can submit a report on a crag's page.
4

Run AI analysis

Everything is assembled into a structured prompt and sent to Anthropic's Claude. The model reasons through four areas before forming a verdict.

Moisture state
Current wetness from recent rain & dry days
Drying analysis
How fast this site dries given exposure & forecast
Structural risk
Hold failure & rock damage risk, especially for sandstone
Seasonal factors
Sun angle, humidity baseline, evaporation rates
5

Return a verdict

The model returns a structured response: verdict, confidence score, summary, safety warnings, contributing factors, recommendations, and a 5-day outlook.

Likely Safe to Climb Marginal — Assess Conditions Do Not Climb Do Not Climb
More detail
You can also ask follow-up questions in context — share what you can see at the crag, describe local micro-conditions, or ask about specific routes. The AI has the full analysis and all previous follow-ups in view when it responds.

Community input

The analysis gets better with real-world feedback. Every crag page has tools for community contribution.

Condition reports

Report what you see: dry, damp, wet, seeping. Add photos for extra context.

Local knowledge

Share seepage lines, drainage quirks, or micro-conditions that only locals know.

Inconsistency reports

Flag when a crag detail is wrong or outdated — e.g. "Kyloe is no longer sheltered". Helps keep the data accurate.

Verdict opinions

Rate how accurate a verdict was. Aggregate ratings help calibrate the model over time.

Limitations

Drying times vary with rock type, seepage lines, season, and micro-topography. The thresholds used (e.g. 3+ consecutive dry days for sandstone) are based on local climbing ethics guidance and BMC advice, not controlled scientific study.

The AI cannot detect seepage, shade from tree cover, or drainage patterns that aren't encoded in the crag attributes. Always combine this with your own judgement on the day.

Weather conditions only. This tool does not account for seasonal bird nesting restrictions, access agreements, land management closures, or other non-weather factors that may affect whether you should visit a crag. Always check local access information separately.

Running cost

Last 7 days

1,137,161 tokens — £26.42

176 requests

All time

1,137,161 tokens — £26.42

176 requests

Based on Claude Opus 4.6 pricing ($15/M input, $75/M output) at ~0.79 USD/GBP.

Get involved

CountyEthics is a personal project and improves with community input.

Submit reports & local knowledge

Use the tools on each crag page to share condition reports, local tips, or flag inaccurate verdicts.

Request new crags

If your local crag isn't listed, use the request form on the crags page.

Get in touch

Got ideas, feedback, or want to help out? Reach out on Instagram @alexwolf.uk.