Token导航 LogoToken导航

Agent Skill · 研究检索

parking(parking 搜索)

parking 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

parking 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为需要联网;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 19,249。

总安装 5,692周安装 242GitHub Stars 公开资料未说明下载 1,994核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:parking(parking 搜索)
来源仓库:https://github.com/ivangdavila/parking
安装命令:
openclaw skills install parking
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install parking

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

通过提供商选择、实时信号分类以及最喜欢的地点和城市的本地记忆,在全球范围内查找、验证和预订停车位。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install parking;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

parking 怎么安装?

在终端执行 openclaw skills install parking。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

parking 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

parking 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

parking 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

parking 用的人多吗?

公开数据是累计安装 5,692 次,近一周安装 242 次,下载 1,994 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

parking 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
parking

parking 搜索 · 本页

5,692—需要联网MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
Parking Radar
slug
parking
version
1.0.0
homepage
https://clawic.com/skills/parking
description
Find, verify, and reserve parking worldwide with provider selection, live-signal triage, and local memory for favorite places and cities.
changelog
Initial release with global provider coverage, booking workflows, API notes, and local memory for favorite parking habits and discoveries.
metadata
{"clawdbot":{"emoji":"PARK","requires":{"bins":[],"config":["~/parking/"]},"os":["linux","darwin","win32"],"configPaths":["~/parking/"]}}

When to Use

User needs parking now, wants to pre-book for an airport or event, needs the right local parking app for a city, or wants help comparing parking options across providers.

Use this skill when the agent should distinguish maps search from real reservations, choose the right operator or marketplace by country and city, track favorite parking spots, and remember hard-won parking discoveries for later.

Architecture

Memory lives in ~/parking/. If ~/parking/ does not exist, run setup.md. See memory-template.md for structure.

~/parking/
├── memory.md          # Activation defaults, home base, vehicle profile, and booking boundaries
├── favorites.md       # User-approved recurring car parks, venues, and trusted entrances
├── cities.md          # City-by-city provider choices, fallback order, and local quirks
├── findings.md        # User-confirmed discoveries, restrictions, scams, and quality signals
├── sessions/
│   └── YYYY-MM-DD.md  # Complex search trail when a trip or event needs multiple comparisons
└── archive/           # Older notes that should not stay in the hot path

Quick Reference

Load only the smallest file needed for the current parking job.

TopicFile
Setup and activation behaviorsetup.md
Memory schema and local note templatesmemory-template.md
Global provider and market matrixprovider-registry.md
Major city patterns and fallback ordercity-patterns.md
Short workflows for search, booking, and verificationexecution-playbook.md
Public and partner API notesapi-notes.md
Official research source ledgersource-ledger.md

Requirements

  • No credentials are required for planning, provider selection, or official web research.
  • Live bookings, plate entry, or payment flows require the user to approve the chosen provider and any data sent to it.
  • Ask before creating persistent local parking memory. Offer a one-off mode when the user does not want favorites, city defaults, or discovery notes saved.
  • Do not promise a guaranteed space unless the provider explicitly exposes reservation inventory or confirmed booking status.

Coverage

This skill is built for the parking cases where generic map search is not enough:

  • quick parking nearby right now
  • airport, train-station, stadium, and venue parking that benefits from pre-booking
  • repeated home-city parking where favorites, vehicle limits, and local operator habits matter
  • travel planning where the correct provider changes by country or even by city
  • parking platform research and API scouting for product or integration work

Core Rules

1. Classify the parking mode before searching

  • Separate the job into one of five modes: find-now, pre-book, recurring-city, operator research, or API scouting.
  • Parking answers get worse fast when the agent mixes instant parking, on-street payment, and guaranteed reservation into one workflow.

2. Default to the lightest successful path

  • If the user just needs parking now, start with map discovery plus the strongest official operator or aggregator for that city.
  • Use deeper provider comparison only when the quick path is ambiguous, expensive, full, or clearly unreliable.
  • Do not turn a simple "find me parking nearby" request into a long market audit.

3. Distinguish discovery, payment, and reservation surfaces

  • Google Maps, Apple Maps, and Parkopedia are discovery surfaces first.
  • Apps such as RingGo, PayByPhone, EasyPark, APCOA Connect, and Parking.sg often solve payment or session start, not guaranteed inventory.
  • Reservation marketplaces such as SpotHero, ParkWhiz, JustPark EventPass, Onepark, Parclick, Q-Park Prebook, and many operator-native airport or venue flows are the right surfaces for guaranteed entry when available.

4. Verify reservation-critical constraints before recommending

  • Check entry window, exit rules, in-and-out privileges, height limit, EV charging, motorcycle or van support, license-plate requirements, and cancellation terms.
  • For airports and events, confirm walking distance, shuttle dependency, venue match, and whether the booking is for a specific car park or only a zone.
  • A cheap parking option is wrong if the user's vehicle or timing cannot actually use it.

5. Be precise about availability language

  • "Reservable now" means the provider is offering a booking path for the requested window.
  • "Reported open" means a listing says the facility is operating, but not that a bay is guaranteed.
  • "Real-time occupancy" should only be claimed when an official provider or city API exposes it directly.

6. Keep a local parking memory only with consent

  • Offer to remember home city, recurring venues, favorite facilities, vehicle constraints, and trusted providers locally in ~/parking/.
  • Store durable notes, not raw payment data or full booking receipts.
  • Promote user-confirmed discoveries such as hidden entrances, repeat scams, and reliable operator patterns into the local favorites, cities, or findings notes.

7. Treat city and country as first-class routing signals

  • Use provider-registry.md to choose the right platform family for the market before assuming Google results or one familiar app will work.
  • Dense urban centers often use payment apps and operator-native garages, while airports and major venues often use reservation marketplaces.
  • When entering a new market, verify the dominant local stack instead of assuming North American or UK patterns apply worldwide.

8. Keep research evidence attached to local discoveries

  • Save only discoveries that came from the user, an official provider, or a clearly identified city or operator source.
  • Tag low-confidence hearsay as unverified and keep it out of favorites until it repeats or gets confirmed.
  • Use source-ledger.md and api-notes.md to separate facts, partnership-only APIs, and "no public API found" cases.

Common Traps

  • Treating a pay-by-phone parking app as proof that a space can be reserved -> many apps start a session only after the driver finds a bay.
  • Claiming "spots available" from a static listing page -> open hours and active inventory are not the same thing.
  • Ignoring plate, height, or vehicle-type restrictions -> the booking looks fine until gate entry fails.
  • Recommending street parking for a high-stress event or airport drop -> uncertainty and enforcement risk overwhelm the small savings.
  • Assuming one global app dominates every market -> parking is highly local and often city-specific.
  • Saving raw booking confirmations or payment details -> keep reusable parking knowledge, not sensitive receipts.

External Endpoints

EndpointData SentPurpose
https://maps.google.com and https://maps.apple.complace text, coordinates, route contextQuick parking discovery, map links, and local navigation handoff
https://www.parkopedia.com and Parkopedia parking data productslocation, time window, optional vehicle filtersGlobal parking discovery, facility detail, and broad market coverage
https://spothero.com and https://api.parkwhiz.comdestination, venue, dates, time window, optional vehicle metadataNorth America reservation search and parking booking APIs
https://www.justpark.com, https://eventpass.justpark.com, https://ringgo.co.uk, https://www.q-park.co.uk, https://www.apcoaconnect.com, https://www.paybyphone.comlocation code, destination, plate, and time windowUK and Europe parking reservation or payment workflows
https://parclick.com, https://www.onepark.co, https://parkmobile.io, https://parkplus.io, https://parking.sg, and official city data portals such as https://api.data.gov.hkdestination, date range, vehicle type, plate when required, and occupancy query parametersRegional booking, payment, or official parking data lookups

No other data should be sent externally unless the user approves another official provider or city source.

Security & Privacy

Data that may leave your machine:

  • parking destination or venue
  • city and time window
  • coordinates and route context
  • plate number only when the chosen provider requires it for entry or billing
  • vehicle profile fields such as height or EV requirement when they change the result

Data that stays local:

  • favorite parking places and entrances in the local favorites note
  • home-city defaults and provider order in the local city notes
  • user-confirmed discoveries and warnings in the local findings note
  • activation rules and parking boundaries in ~/parking/memory.md

This skill does NOT:

  • store payment card data, account passwords, or unnecessary booking receipts
  • claim guaranteed inventory without provider evidence
  • use undeclared anti-bot scraping flows
  • modify its own SKILL.md

Trust

This skill can send parking queries, dates, route context, plate numbers, or vehicle constraints to the selected parking marketplace, operator, payment app, maps provider, or official city data source. Only use live parking flows if you trust the chosen provider with that data.

Scope

This skill ONLY:

  • finds, compares, and explains parking options
  • chooses the right provider family for each city and task
  • verifies reservation-critical constraints before the user commits
  • maintains lightweight local memory for favorites, cities, and discoveries after approval

This skill NEVER:

  • invent live availability or occupancy
  • assume a payment app guarantees entry
  • treat one country's parking workflow as universal
  • persist sensitive payment or identity data without a clear need

Related Skills

Install with clawhub install <slug> if user confirms:

  • maps - Use map providers, geocoding, and route links to ground parking search in real geography.
  • travel - Fold parking choices into broader trip timing, lodging, and movement decisions.
  • booking - Compare reservation terms, cancellation rules, and booking friction across providers.
  • compare - Turn multiple parking options into a clear side-by-side decision.
  • car-rental - Add pickup, return, and vehicle-access context when parking decisions sit inside a driving trip.

Feedback

  • If useful: clawhub star parking
  • Stay updated: clawhub sync

平台分布

OpenClaw

93.46%
按下载量换算1,864

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills