farming-expert — скилл для Claude Code: что делает и как установить
Скилл farming-expert даёт агенту агрономический контекст: севооборот, показания почвенных датчиков, NDVI и прогноз урожайности, с примерами кода на Python.
Apache-2.0 283 установок ★ 45 обновлён 2026-10-12
npx skills add personamanagmentlayer/pcl --skill farming-expert -a claude-codeЧто делает farming-expert
Скилл добавляет агенту предметную область: точное земледелие, управление хозяйством, мониторинг посевов и агротех. Внутри есть справочная часть про GPS-технику, дифференцированное внесение, анализ почвы и снимки с дронов или спутников. Рядом лежит код на Python: класс FarmManagementSystem планирует севооборот, считает здоровье культуры по датчикам и оценивает урожайность поля.
Второй пример, AgricultureIoT, усредняет свежие показания почвенного зонда: влажность, температуру, pH, электропроводность и питательные элементы. Дальше он помечает аномалии. Пороги стоят прямо в коде: влажность ниже 20, pH вне 5.5–7.5, NDVI ниже 0.6.
Код ждёт готовый слой доступа к данным: объект db с методами вроде get_field и get_crop_history. В примерах вызывается np.mean, поэтому без numpy они не запустятся. Скиллу разрешены только Read, Write и Edit — файлы он правит, а команды выполняете вы.
Какие задачи решает
- спланировать севооборот на несколько лет с пояснением по каждой культуре
- собрать мониторинг поля по датчикам и NDVI
- прикинуть урожайность участка по площади, осадкам и внесённым удобрениям
- разобрать показания почвенного зонда и поймать выход pH за норму
- описать поле и цикл культуры структурами данных в Python
Примеры запросов агенту
- Спланируй севооборот для поля
field-12на три года и объясни выбор каждой культуры - Напиши обработчик почвенных датчиков: усредни влажность и pH за сутки и выдай алерт, если влажность ниже 20
- Оцени урожайность поля по NDVI и накопленным осадкам
Как установить farming-expert
1Через Skills CLI (Vercel)
npx skills add personamanagmentlayer/pcl --skill farming-expert -a claude-code2Вручную: положить файлы
~/.claude/skills/farming-expert/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/farming-expert/
1Через Skills CLI (Vercel)
npx skills add personamanagmentlayer/pcl --skill farming-expert -a codex2Вручную: положить файлы
.agents/skills/farming-expert/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add personamanagmentlayer/pcl --skill farming-expert -a cursor2Вручную: положить файлы
.cursor/skills/farming-expert/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install personamanagmentlayer/pcl stdlib/domains/farming-expert2Вручную: положить файлы
.github/skills/farming-expert/SKILL.md1Через Skills CLI (Vercel)
npx skills add personamanagmentlayer/pcl --skill farming-expert -a gemini-cli2Командой платформы
gemini skills install https://github.com/personamanagmentlayer/pcl --path stdlib/domains/farming-expertВручную: положить файлы
.devin/skills/farming-expert/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Через Skills CLI (Vercel)
npx skills add personamanagmentlayer/pcl --skill farming-expert -a opencodeClaude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о farming-expert
Как установить farming-expert в Claude Code?
Выполните npx skills add personamanagmentlayer/pcl --skill farming-expert -a claude-code. После этого файл скилла лежит в ~/.claude/skills/farming-expert/SKILL.md.
Что делает farming-expert?
Подсказывает агенту, как устроены севооборот, мониторинг посевов и прогноз урожайности, и даёт готовые классы на Python для расчётов по полю и датчикам.
Что нужно, чтобы примеры кода заработали?
Нужен numpy для np.mean и объект db с методами вроде get_field и get_crop_history — скилл опирается на уже существующий слой данных.
Исходный текст
SKILL.md · Apache-2.0
---
name: farming-expert
version: 2.0.0
description: >-
Expert-level precision agriculture, farm management systems, crop monitoring, and agtech.
Use when the user mentions agriculture, precision agriculture, agtech, or crop
management, or when the task involves Farm Management or AgTech Solutions.
category: domains
tags: [agriculture, farming, precision-agriculture, agtech, crop-management]
allowed-tools:
- Read
- Write
- Edit
---
# Farming Expert
Expert guidance for precision agriculture, farm management systems, crop monitoring, IoT sensors, and agricultural technology.
## Core Concepts
### Precision Agriculture
- GPS-guided equipment
- Variable rate technology
- Crop monitoring and sensors
- Soil analysis and mapping
- Drone/satellite imagery
- Automated irrigation systems
### Farm Management
- Crop planning and rotation
- Resource optimization
- Yield prediction
- Weather forecasting integration
- Equipment maintenance
- Financial management
### AgTech Solutions
- IoT sensors (soil, weather)
- Machine learning for yield prediction
- Automated harvesting
- Livestock tracking
- Supply chain integration
- Marketplace platforms
## Farm Management System
```python
from dataclasses import dataclass
from typing import List, Optional
from datetime import datetime, timedelta
from enum import Enum
class CropType(Enum):
WHEAT = "wheat"
CORN = "corn"
SOYBEANS = "soybeans"
RICE = "rice"
VEGETABLES = "vegetables"
class GrowthStage(Enum):
PLANTED = "planted"
GERMINATION = "germination"
VEGETATIVE = "vegetative"
FLOWERING = "flowering"
HARVEST_READY = "harvest_ready"
HARVESTED = "harvested"
@dataclass
class Field:
field_id: str
name: str
area_hectares: float
soil_type: str
coordinates: List[tuple] # GPS polygon
irrigation_system: str
drainage_quality: str
@dataclass
class CropCycle:
cycle_id: str
field_id: str
crop_type: CropType
variety: str
planting_date: datetime
expected_harvest_date: datetime
growth_stage: GrowthStage
seed_rate: float
fertilizer_applied: List[dict]
pesticides_applied: List[dict]
irrigation_schedule: List[dict]
class FarmManagementSystem:
"""Farm management and crop tracking"""
def __init__(self, db):
self.db = db
def plan_crop_rotation(self, field_id, years=3):
"""Generate crop rotation plan"""
field = self.db.get_field(field_id)
history = self.db.get_crop_history(field_id, years=10)
# Analyze soil nutrients and previous crops
rotation_plan = []
# Rules for rotation:
# - Alternate nitrogen-fixing and nitrogen-demanding crops
# - Avoid same crop family consecutively
# - Consider soil health and pest management
for year in range(years):
recommended_crop = self.recommend_next_crop(field, history, year)
rotation_plan.append({
'year': datetime.now().year + year,
'crop': recommended_crop,
'reason': self.explain_recommendation(recommended_crop, history)
})
return rotation_plan
def monitor_crop_health(self, field_id):
"""Monitor crop health using sensor data"""
field = self.db.get_field(field_id)
current_crop = self.db.get_current_crop(field_id)
# Collect sensor data
soil_moisture = self.get_soil_moisture_data(field_id)
weather_data = self.get_weather_data(field.coordinates)
ndvi_data = self.get_ndvi_from_satellite(field.coordinates)
# Analyze health indicators
health_score = self.calculate_health_score(
soil_moisture,
weather_data,
ndvi_data,
current_crop
)
alerts = []
if soil_moisture < current_crop.optimal_moisture_min:
alerts.append({
'type': 'irrigation_needed',
'severity': 'high',
'message': 'Soil moisture below optimal level'
})
if ndvi_data < 0.6: # Vegetation health threshold
alerts.append({
'type': 'crop_stress',
'severity': 'medium',
'message': 'NDVI indicates possible crop stress'
})
return {
'field_id': field_id,
'health_score': health_score,
'soil_moisture': soil_moisture,
'ndvi': ndvi_data,
'alerts': alerts,
'recommendations': self.generate_recommendations(alerts)
}
def predict_yield(self, field_id):
"""Predict crop yield using ML"""
field = self.db.get_field(field_id)
current_crop = self.db.get_current_crop(field_id)
# Features for prediction
features = {
'field_area': field.area_hectares,
'soil_type': field.soil_type,
'crop_variety': current_crop.variety,
'days_since_planting': (datetime.now() - current_crop.planting_date).days,
'total_rainfall': self.get_accumulated_rainfall(field_id),
'avg_temperature': self.get_avg_temperature(field_id),
'fertilizer_amount': sum(f['amount'] for f in current_crop.fertilizer_applied),
'ndvi_avg': self.get_avg_ndvi(field_id)
}
# Use trained model to predict yield
predicted_yield_per_hectare = self.yield_model.predict([features])[0]
total_yield = predicted_yield_per_hectare * field.area_hectares
return {
'field_id': field_id,
'predicted_yield_kg': total_yield,
'yield_per_hectare': predicted_yield_per_hectare,
'confidence': 0.85,
'expected_harvest_date': current_crop.expected_harvest_date
}
```
## IoT Sensor Integration
```python
class AgricultureIoT:
"""IoT sensor data collection and analysis"""
def process_soil_sensor_data(self, sensor_id):
"""Process soil sensor readings"""
readings = self.db.get_recent_readings(sensor_id, hours=24)
analysis = {
'sensor_id': sensor_id,
'avg_moisture': np.mean([r['moisture'] for r in readings]),
'avg_temperature': np.mean([r['temperature'] for r in readings]),
'avg_ph': np.mean([r['ph'] for r in readings]),
'avg_ec': np.mean([r['ec'] for r in readings]), # Electrical conductivity
'nitrogen_level': np.mean([r['nitrogen'] for r in readings]),
'phosphorus_level': np.mean([r['phosphorus'] for r in readings]),
'potassium_level': np.mean([r['potassium'] for r in readings])
}
# Detect anomalies
anomalies = []
if analysis['avg_moisture'] < 20:
anomalies.append('Low soil moisture - irrigation recommended')
if analysis['avg_ph'] < 5.5 or analysis['avg_ph'] > 7.5:
anomalies.append(f'Soil pH out of optimal range: {analysis["avg_ph"]:.1f}')
analysis['anomalies'] = anomalies
return analysis
def automate_irrigation(self, field_id):
"""Automated irrigation control"""
field = self.db.get_field(field_id)
soil_moisture = self.get_soil_moisture_data(field_id)
weather_forecast = self.get_weather_forecast(field.coordinates, days=3)
# Decision logic
should_irrigate = False
duration_minutes = 0
# Check if irrigation is needed
if soil_moisture < field.moisture_threshold:
# Check if rain is expected
expected_rainfall = sum(day['rainfall_mm'] for day in weather_forecast)
if expected_rainfall < 10: # Less than 10mm expected
should_irrigate = True
# Calculate irrigation duration
moisture_deficit = field.moisture_threshold - soil_moisture
duration_minutes = int(moisture_deficit * field.area_hectares * 60 / field.irrigation_rate)
if should_irrigate:
self.activate_irrigation(field_id, duration_minutes)
return {
'field_id': field_id,
'irrigation_activated': should_irrigate,
'duration_minutes': duration_minutes,
'reason': 'Soil moisture below threshold' if should_irrigate else 'No irrigation needed'
}
```
## Weather and Climate Analysis
```python
class WeatherAnalytics:
"""Weather-based agricultural decisions"""
def analyze_growing_conditions(self, field_id, date_range):
"""Analyze weather suitability for crops"""
weather_data = self.get_historical_weather(field_id, date_range)
# Calculate growing degree days (GDD)
gdd = sum([
max(0, (day['temp_max'] + day['temp_min']) / 2 - 10)
for day in weather_data
])
# Analyze frost risk
frost_days = len([d for d in weather_data if d['temp_min'] < 0])
# Water balance
total_rainfall = sum(d['rainfall_mm'] for d in weather_data)
total_evapotranspiration = sum(d['et_mm'] for d in weather_data)
water_deficit = total_evapotranspiration - total_rainfall
return {
'growing_degree_days': gdd,
'frost_days': frost_days,
'total_rainfall_mm': total_rainfall,
'water_deficit_mm': water_deficit,
'avg_temperature': np.mean([d['temp_avg'] for d in weather_data]),
'suitability_score': self.calculate_suitability_score(gdd, frost_days, water_deficit)
}
def optimize_planting_date(self, field_id, crop_type):
"""Determine optimal planting date"""
historical_weather = self.get_historical_weather(field_id, years=10)
crop_requirements = self.get_crop_requirements(crop_type)
# Find window with:
# - Minimal frost risk
# - Adequate soil temperature
# - Good rainfall distribution
optimal_dates = []
for year_data in historical_weather:
for date, conditions in year_data.items():
score = self.score_planting_conditions(
conditions,
crop_requirements
)
optimal_dates.append((date, score))
# Return best planting window
best_dates = sorted(optimal_dates, key=lambda x: x[1], reverse=True)[:10]
return {
'recommended_planting_window': {
'start': best_dates[-1][0],
'end': best_dates[0][0]
},
'confidence': np.mean([d[1] for d in best_dates])
}
```
## Pest and Disease Management
```python
class PestManagement:
"""Pest and disease monitoring and management"""
def detect_pest_risk(self, field_id):
"""Predict pest pressure"""
weather_data = self.get_recent_weather(field_id, days=14)
crop = self.db.get_current_crop(field_id)
# Environmental factors affecting pests
avg_temp = np.mean([d['temperature'] for d in weather_data])
avg_humidity = np.mean([d['humidity'] for d in weather_data])
rainfall = sum(d['rainfall_mm'] for d in weather_data)
risk_factors = {
'temperature_risk': self.assess_temp_risk(avg_temp, crop.crop_type),
'humidity_risk': self.assess_humidity_risk(avg_humidity),
'rainfall_risk': self.assess_rainfall_risk(rainfall)
}
overall_risk = sum(risk_factors.values()) / len(risk_factors)
recommendations = []
if overall_risk > 0.7:
recommendations.append('Scout fields for pest activity')
recommendations.append('Consider preventive treatment')
elif overall_risk > 0.5:
recommendations.append('Increase monitoring frequency')
return {
'field_id': field_id,
'overall_risk_score': overall_risk,
'risk_level': 'high' if overall_risk > 0.7 else 'medium' if overall_risk > 0.4 else 'low',
'risk_factors': risk_factors,
'recommendations': recommendations
}
def analyze_crop_image(self, image_data):
"""Detect diseases from crop images using ML"""
# Would use computer vision model (CNN) trained on crop diseases
# Returns detected diseases and confidence scores
predictions = self.disease_detection_model.predict(image_data)
return {
'detected_diseases': predictions['diseases'],
'confidence': predictions['confidence'],
'affected_area_percent': predictions['affected_area'],
'treatment_recommendations': self.get_treatment_plan(predictions['diseases'])
}
```
## Best Practices
- Use precision agriculture techniques
- Implement crop rotation
- Monitor soil health regularly
- Integrate weather data for decisions
- Use IoT sensors for real-time monitoring
- Apply variable rate technology
- Optimize water usage
- Practice integrated pest management
- Track field-level profitability
- Use data-driven decision making
- Maintain equipment properly
- Follow sustainable practices
## Anti-Patterns
❌ Over-application of inputs
❌ Ignoring soil health
❌ No crop rotation
❌ Manual data collection only
❌ Ignoring weather forecasts
❌ Reactive instead of proactive management
❌ No yield analysis
## Resources
- Precision Agriculture: https://www.ispag.org/
- FAO Agricultural Data: http://www.fao.org/faostat/
- Crop Science Society: https://www.crops.org/
- Agricultural IoT: https://www.iot-agriculturet.com/
© personamanagmentlayer · Apache-2.0 · personamanagmentlayer/pcl