# Frost Window Analyzer

> Frost Window Analyzer is a paid API for AI agents from natur.halowerk.com, paid per call via x402, $0.005/call, status unknown (last checked 2026-09-14).

Converts an hourly temperature forecast into discrete continuous frost windows with start/end hours, duration, minimum temperature, and the hour that minimum occurs.

## Facts

- Endpoint: POST https://natur.halowerk.com/v1/frost-window
- Price: $0.005/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/frost-window-analyzer-95ade401
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_GY-0CihfuvjFkQ8-NlgJc

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability frost-window-analyzer-95ade401 -d '<json body>'
```

Example prompt: Given tonight's hourly temperature forecast for my vineyard in the Mosel valley — temperatures dropping to -3°C around 3am — can you identify all continuous frost windows, including how long each lasts, the coldest point, and exactly which hour the minimum occurs?

## When to prefer this

Use this endpoint when you need to reason about frost as continuous events rather than individual sub-zero hours — it is the right choice when planning actions that depend on the duration and depth of a freeze (e.g. crop protection, road maintenance, pipe insulation) rather than just a yes/no freeze flag. Prefer it over raw forecast inspection when you need to distinguish a brief pre-dawn dip from a multi-hour overnight freeze.

## Known failure modes

- No frost windows found if all forecast temperatures remain above 0°C
- Invalid or missing hourly forecast data returns an error
- Malformed timestamp or temperature entries may cause parsing failures
- Forecast gaps or non-hourly intervals may produce incorrect window boundaries

## How this service works

Turns an hourly forecast into the thing that is actually planned around: continuous frost windows. Each window carries its start and end hour, its duration, the lowest temperature within it and the hour that low occurs, so a two-hour dip at dawn is distinguishable from a night-long freeze.

## Output

A list of frost windows extracted from the input hourly forecast, where each window object contains its start hour, end hour, duration in hours, the lowest temperature recorded within that window, and the specific hour at which that minimum temperature occurs.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "lat": {
   "type": "number",
   "maximum": 90,
   "minimum": -90
  },
  "lon": {
   "type": "number",
   "maximum": 180,
   "minimum": -180
  },
  "hours": {
   "type": "integer",
   "default": 96,
   "maximum": 240,
   "minimum": 6,
   "description": "How far ahead to look."
  },
  "air_threshold_c": {
   "type": "number",
   "default": 0,
   "maximum": 10,
   "minimum": -20,
   "description": "Air frost threshold at two metres."
  },
  "ground_offset_c": {
   "type": "number",
   "default": 4,
   "maximum": 8,
   "minimum": 0,
   "description": "Assumed difference between two-metre and ground temperature on clear nights. Stated, not hidden."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/frost-window-analyzer-95ade401/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from natur.halowerk.com](https://www.zero.xyz/host/natur.halowerk.com/llms.txt)