安装指南

1. ClawHub CLI 从源直接安装技能的最快方式。
   npx clawhub@latest install houseplant-watering-computer
2. 手动安装 将技能文件夹复制到以下位置之一
   全局模式
   ~/.openclaw/skills/
   工作区
   /skills/
   优先级：工作区 > 本地 > 内置
3. 提示词安装 将此提示词复制到 OpenClaw 即可自动安装。
   请帮我使用 Clawhub 安装 houseplant-watering-computer。如果尚未安装 Clawhub，请先安装（npm i -g clawhub）。
   更倾向于下载？
   获取包含原生技能文件的ZIP压缩包。
   下载技能 (.zip)

什么是

Houseplant Watering Computer is an offline Python skill that turns pot and room details into practical, per-plant watering guidance. It estimates substrate water-holding capacity, pot volume, plant demand, and environmental effects such as temperature and humidity to calculate a usable water reservoir, watering dose, and expected dryout time.
Use it to replace guesswork with moisture checks, interpret signs of overwatering or underwatering, and plan self-watering or cache-pot reservoirs. As a focused addition to Openclaw Skills, it helps tailor care to each pot and changing conditions while emphasizing that calculated intervals are prompts to check moisture—not a substitute for checking the soil.

应用场景

Caring for a new plant:
Estimate a starting watering dose and moisture-check interval from its substrate, pot dimensions, plant size, and light.
Troubleshooting recurring plant decline:
Route yellowing, wilting despite wet soil, or fungus gnats toward likely overwatering, underwatering, or other causes.
Adapting care to seasonal changes:
Account for room temperature and humidity when winter heating or summer heat changes plant water demand.
After repotting:
Recalculate when pot size or substrate mix changes the soil's water-holding capacity and dryout time.
Planning time away:
Estimate reservoir size and refill cadence for suitable wick or cache-pot setups.
Checking before watering:
Interpret finger-test depth, soil feel, or moisture observations and determine when to check again.

工作原理

Inventory the pot: Record substrate mix, top and bottom pot diameters, pot height, plant size class, light level, room temperature, and humidity.
Compute the water balance: Estimate substrate water-holding capacity and pot volume, using tapered-pot or cylinder geometry as appropriate.
Estimate plant demand: Use plant size as a leaf-area proxy, alongside light and vapor pressure deficit derived from temperature and relative humidity.
Get actionable output: Review the estimated dryout time, usable water reservoir in milliliters, and per-watering dose.
Check actual moisture: Run check with observed finger-test depth and soil feel to get a wet/dry verdict and suggested next-check timing.
Investigate symptoms or reservoirs: Use symptoms to assess common watering problems and fungus gnats, or selfwatering to size a reservoir and estimate fill cadence.
Verify before acting: Check soil moisture before watering, especially for intervals longer than four days, and recalculate after changes in season, substrate, or pot.

配置指南

The skill requires Python 3 and uses only the standard library; it works offline. Run the script from the skill directory. Supply measurements and conditions that reflect the actual pot and growing environment.

Estimate watering guidance for a Monstera in a tapered pot

python3 scripts/watering_computer.py compute --substrate peat-mix --pot-tapered 24,18,22
--plant large-foliage --light bright-indirect --temp 22 --humidity 45

Interpret a shallow finger test when the soil feels dry

python3 scripts/watering_computer.py check --finger-depth shallow --feels dry --pot-tapered 24,18,22

Assess lower-leaf yellowing, wet soil, and fungus gnats

python3 scripts/watering_computer.py symptoms --yellowing lower --soil wet --gnats

Estimate self-watering reservoir needs for a 21-day absence

python3 scripts/watering_computer.py selfwatering --pot-tapered 24,18,22 --substrate peat-mix --away-days 21
Measure tapered pots at the top and bottom and include height. Identify the mix honestly, assess light where the plant sits, and provide room temperature in Celsius and relative humidity. Treat calculated schedules as estimates: check moisture before watering.
Houseplant Watering Computer 数据架构与分类体系
The skill uses per-pot inputs and produces care guidance; the supplied SKILL.md does not specify a persistent data store or generated file format.
Input or output Details
Substrate Named mix, such as peat-mix; substrate properties inform water-holding capacity (WHC).
Pot geometry Tapered pot dimensions as top diameter, bottom diameter, and height; volume is estimated from tapered-cone geometry. Cylinder geometry is also supported by the method.
Plant and environment Plant size class, light level, temperature, and relative humidity; temperature and humidity inform vapor pressure deficit (VPD).
compute results Estimated dryout time, usable water reservoir in mL, and watering dose in mL.
check observations Finger-test depth and soil feel or moisture observations; used to interpret wet/dry status and next-check timing.
symptoms observations Visual symptoms and soil or fungus-gnat observations used for diagnostic routing.
selfwatering inputs Pot, substrate, and days away; used to estimate reservoir sizing and fill cadence.
Reference material is organized in references/water-balance.md for WHC, pot geometry, VPD, plant demand, and dryout calculations, and references/symptom-router.md for symptom routing, fungus-gnat guidance, and self-watering or cache-pot engineering. The skill metadata identifies the name, version, author, license, and tags.

高级特性

Per-pot physical estimates:
Combines substrate WHC, pot geometry, plant-size demand, light, and VPD rather than applying a universal watering calendar.
Tapered-pot geometry:
Accounts for top and bottom diameters and height instead of treating every container as a cylinder.
Environment-sensitive demand:
Temperature and relative humidity modify estimated water use; the documented examples range from a 0.6 multiplier in cool, humid conditions to 1.6 in hot, dry conditions.
Multiple diagnostic workflows:
Separate
compute
,
check
, and
symptoms
commands cover planning, current moisture observations, and symptom interpretation.
Self-watering and cache-pot planning:
Estimates reservoir sizing and fill cadence while flagging always-wet substrate and rot-prone plant concerns.
Offline, dependency-light operation:
Runs with Python 3 standard library only, without requiring a network connection.
Practical verification guidance:
Encourages moisture checks before watering and highlights risks from oversized pots, tiny splashes, and schedule-driven care.

name: houseplant-watering-computer description: Use when houseplants keep dying or schedules feel like guesswork — computes real soil water balance per pot (substrate water-holding capacity, pot volume, plant size leaf area, light level, temperature/VPD, humidity), tells you HOW MUCH and HOW OFTEN to water, detects over- vs under-watering symptoms, handles self-watering pots and cache pots, and generates a per-pot care card with moisture-check dates. version: 1.0.0
author: Denis Voronin
license: MIT
tags: [houseplants, watering, plants, soil-moisture, gardening, plant-care]

Houseplant Watering Computer

"Water weekly" is why houseplants die. The right question is: how fast does THIS pot dry out? — a function of substrate water-holding capacity, pot volume, plant size, light, temperature, and humidity. This skill computes the soil water balance per pot and turns it into an actionable answer: how much water, how often, and when to check next.

When to Use

New plant, no idea how often to water it
Repeated plant deaths (root rot = overwatering is the #1 killer)
Changing seasons: the winter schedule isn't the summer schedule
Repotting: substrate change changes everything

Setting up self-watering pots / cache pots correctly

Symptoms: yellowing, wilting despite wet soil, fungus gnats

Don't use for: hydroponics (different physics), outdoor gardens (rain dominates — use ET-based scheduling), lawns.

The Method

Inventory each pot: substrate mix, pot dimensions (top/bottom diameter + height), plant size class, light level, room temp/humidity.
compute — per-pot physics:
Substrate WHC (water-holding capacity) L per L of soil from the mix
Pot volume from truncated-cone geometry (tapered pots) or cylinder
Plant water use from leaf area proxy + light + VPD (vapor pressure deficit from temp/RH)
→ Dryout time: days from saturation to the plant's wilt/watering point, with the usable reservoir in mL and per-watering dose in mL
check — interpret what you observe (finger test depths, moisture meter readings, weight) into wet/dry verdicts and next-check timing.
symptoms — route visual symptoms to over- vs under-watering and other causes; fungus gnat lifecycle breaks.
selfwatering — reservoir sizing and fill cadence for wick/cache pots; when NOT to use them (mines: always-wet substrate + rot-prone species).

Quick Start

# Monstera in a 24cm tapered pot, peat-based mix, bright indirect, 22C/45%RH
python3 scripts/watering_computer.py compute --substrate peat-mix --pot-tapered 24,18,22 \

--plant large-foliage --light bright-indirect --temp 22 --humidity 45

# Check what I'm feeling in the pot right now
python3 scripts/watering_computer.py check --finger-depth shallow --feels dry --pot-tapered 24,18,22

# Diagnose: yellow lower leaves + fungus gnats + soil stays wet 10+ days
python3 scripts/watering_computer.py symptoms --yellowing lower --soil wet --gnats

# Self-watering pot reservoir sizing
python3 scripts/watering_computer.py selfwatering --pot-tapered 24,18,22 --substrate peat-mix --away-days 21

Stdlib-only Python 3. Offline.

Substrate Science (what drives everything)
Mix WHC (mL/L) Dryout Best for
Peat/coir-based potting mix 450–550 Medium Most tropicals, foliage
Heavy garden soil in a pot 600+ Very slow Nobody — compacts, rots 50/50 peat + perlite 380–430 Faster Seedlings, succulent-adjacent Succulent/cactus mix 250–320 Fast Sansevieria, succulents, cacti
Bark-heavy aroid mix 180–250 Very fast Philodendron, orchid-adjacent
Semi-hydro (LECA) 300–400 Wick-driven Cache-pot setups

The watering INTERVAL the plant experiences is substrate × pot size × demand — never a calendar. A peat mix in a 12 cm pot dries in ~3 days in summer; the same plant in a 24 cm pot takes 7–9.

Environment Multipliers (VPD)

Demand scales with warmth and dryness of air. The script computes vapor pressure deficit from temp + RH and multiplies plant water use:

18 °C / 70% RH → ×0.6 (winter, slow everything) 22 °C / 45% RH → ×1.0 (baseline) 26 °C / 30% RH → ×1.6 (heatwave, AC, radiator proximity)

Winter heating is the hidden killer: soil stays cold and wet (slow uptake) while air is dry — check less, water less, don't fertilize.

Common Pitfalls

Calendar watering. The #1 killer. Moisture-check before every watering for the first month with any new pot; the calendar emerges from the physics, per pot.
Watering on a schedule that ignores pot size. Same plant, doubled pot volume → interval nearly doubles. Small pots dry in days; big pots hold a hidden reservoir for a week+.
"Drainage fixes overwatering." Drainage holes move water OUT after saturation; they do nothing for a substrate kept continuously wet by too-frequent watering.
Even soak vs splash. A 200 mL splash on a 6 L pot wets the top 2 cm, salts accumulate in the root zone, and the bottom roots stay dry. Water to ~20–25% of substrate volume, until slight runoff.
Cold tap water on tropicals. Let it reach room temperature; some species are sensitive to chlorine — overnight resting resolves both.
Repotting into "bigger is better." Too large a pot = wet soil the roots never reach = rot. Go up ≤ 5 cm in diameter.
Self-watering pots as a cure for underwatering. They fix interval management, not light/substrate problems; rot-prone species need a timer-free setup or a very airy mix.

Verification Checklist

Pot measured (top, bottom diameters, height) — tapered ≠ cylinder
Substrate identified honestly (what's actually in the pot)
Light level honestly assessed (what hits the plant, not the window)
compute output dose is ~15–25% of substrate volume
Finger/weight check performed before acting on any interval > 4 days
Winter: interval stretched, fertilizer stopped

References

references/water-balance.md — the physics: WHC, pot geometry math, VPD tables, plant demand classes, dryout-time derivation
references/symptom-router.md — full visual symptom routing table, fungus gnat protocol, self-watering/cache-pot engineering
密码管理器迁移工具

Mattress Spec Decoder