连锁门店覆盖范围分析

安装指南

1. ClawHub CLI 从源直接安装技能的最快方式。
   npx clawhub@latest install chain-store-footprint-analysis
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 chain-store-footprint-analysis。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

连锁门店覆盖范围分析使用 Crawlora 数据集、实时门店查询器和 Google Maps 商家记录，帮助绘制并比较指定品牌的已观察分店。它支持在明确定义的范围内进行分店计数、地理重叠、设施覆盖和覆盖空白分析。
此工作流强调证据质量：跟踪来源 ID、URL、日期、抓取时间、查询边界和结果上限，同时区分已验证位置与不确定的名称匹配。Openclaw Skills 用户可以更透明地了解已观察到的供给情况，避免将不完整记录误认为完整普查、市场份额、需求估算或选址盈利能力评估。

应用场景

比较所选连锁品牌在城市、州、国家或按坐标定义的区域内观察到的分店数量。
识别竞争品牌之间的地理重叠和表面上的覆盖空白。
查看来源记录所提供的已发布设施、所有权或运营属性。
使用实时门店查询器或 Google Maps 地点记录刷新选定位置。
记录不完整的搜索、结果上限、不确定的品牌关联，以及需要进一步核实的区域。

工作原理

定义品牌及其别名、位置范围（包括门店、售货亭或授权经营网点）、地理边界、设施标准、来源集合和采集预算。
使用相关的已存储 Crawlora 数据集和分面、搜索或附近查询来发现候选分店；针对每个来源使用正确的地理筛选条件和分类体系。
使用实时查询器或地点核验来刷新所选记录。将地理编码失败与成功但未返回分店的搜索分别记录，并标记达到结果上限的情况。
在可能的情况下，按品牌和稳定的门店 ID 对记录进行协调。跨来源匹配时，交叉核实地址和坐标；不要仅凭相似名称就认定存在品牌关联。
对定义区域内的分店去重，保留边界和查询证据，并在报告数量的同时列出来源覆盖情况、已观察属性以及未知或不完整区域。
仅将结果解释为已观察到的供给。不要将缺失记录视为门店关闭或需求未满足的证据，也不要仅凭数量推断市场份额。

配置指南

在环境中设置 Crawlora API 密钥。连接后优先使用列出的 Crawlora MCP 工具；否则使用随附的 scripts/crawlora.sh 客户端。
export CRAWLORA_API_KEY="your-crawlora-api-key"
查阅 reference/endpoints.md 获取准确的 API 方法和参数。随附脚本会向 https://api.crawlora.net/api/v1 发送 x-api-key 标头。
数据集和实时门店查询请求示例：
scripts/crawlora.sh /datasets/starbucks-stores/facets facet=country
scripts/crawlora.sh /datasets/starbucks-stores/search country=US city=Seattle page=1 page_size=100
scripts/crawlora.sh /starbucks/stores 'place=Seattle, WA' market=us
scripts/crawlora.sh /datasets/google-map-businesses/search 'q=Starbucks' city=Seattle page=1 page_size=10
除 HTTP 状态外，还要检查应用级 code；响应内容位于 data 中。遇到身份验证错误时停止，遇到 HTTP 429 时退避重试，瞬时上游故障则重试一次。根据请求范围和额度预算限制页数与调用次数。请求加拿大运营数据时，使用加拿大 Starbucks 主机。

数据架构

位置证据： 将来源 ID、来源 URL、来源日期以及观察或抓取时间保留为单独字段。
地理范围： 保留准确的国家、州、城市、坐标、边界或查询定义，以及请求页数与返回页数的记录。
来源专属元数据： 保持各来源的分类体系和代码相互独立。Starbucks 的 country 表示地理区域，而 market 记录抓取主机来源；Google Maps 的类别和国家/地区值采用不同约定。McDonald's 查询器的市场、结果上限和英里制半径也具有来源专属性。
覆盖范围与限制： 记录搜索结果数、上限标记、result_capped 状态、来源覆盖范围以及未知或不完整区域。
身份与分店属性： 尽可能优先使用门店 ID 对同一品牌的记录去重。保留可交叉核实的地址和坐标证据，并保留有依据的区别，例如搬迁、永久关闭、授权经营或公司直营的位置。
输出： 生成分店证据台账、去重后的区域-品牌矩阵，以及观察到的设施或所有权细分。设施字段为空可能表示未发布相关信息，并不代表该位置没有此设施。
衍生指标： 使用匹配后的唯一分店计数。计算密度前先定义区域；重叠的半径查询需要明确说明其并集边界。人均指标需要单独获取来源可靠、且地理区域和日期相匹配的人口数据。

高级特性

将已存储数据发现与选定的实时查询器或地点核验相结合，以检查信息时效性。
支持多种来源类型，包括 Starbucks 数据集和查询器、Google Maps 商家记录，以及独立的 McDonald's 坐标查询器。
在证据台账中保留来源结果上限、页数、查询边界、抓取时间和不确定性。
考虑来源专属的半径单位和限制；例如，McDonald's 查询器的半径单位为英里，而数据集附近查询的半径单位为米。
区分地理编码失败与成功但未返回分店的搜索，并标记不完整的实时结果。
支持谨慎的跨来源协调；当证据支持将同一综合体内的不同经营单元或售货亭视为独立分店时，可分别处理。
只有在存在可比较的已保存观察记录和经核实的变更时，才支持历史扩张或收缩分析。
Openclaw Skills 报告旨在避免夸大结论：部分样本不是官方普查，搜索结果中缺失不代表门店关闭，覆盖范围数量也无法证明需求、销售额、人流量、市场份额或投资适宜性。

name: chain-store-footprint-analysis description: Compare observed chain-store coverage using Crawlora Starbucks directories, live chain locators, and Google Maps business records. Use for branch counts, geographic overlap, amenity coverage, or footprint gaps within a defined area with source caps and identity uncertainty preserved.

Chain store footprint analysis

Access and scope

Set CRAWLORA_API_KEY to your Crawlora key. Prefer the listed Crawlora MCP tools when connected; otherwise use the bundled scripts/crawlora.sh, which sends x-api-key to https://api.crawlora.net/api/v1. Read reference/endpoints.md for exact methods and parameters. Check application code as well as HTTP status; payloads are inside data. Stop on authentication errors, back off on 429, and retry a transient upstream failure once. Bound calls/pages by the requested scope and credit budget. Retain source IDs, URLs, source dates, and observation/crawl times separately.

Map observed branches for specified chains and geographic boundaries. Establish brands/aliases, store versus kiosk/licensed location scope, cities or coordinate boundaries, amenity criteria, source set, and collection budget before counting. This describes observed supply, not sales, foot traffic, demand, or site profitability.

Discover branches under matched geographic scopes

Use stored Starbucks facets/search/nearby for brand-specific discovery and stable store_number. Dataset country is actual geography; market=us/ca records crawl-host provenance and must not be mistaken for a geographic filter. Preserve exact country/state/city and coordinate evidence.
Refresh selected locations with live Starbucks locator/nearest-store calls. Either place or a coordinate pair is needed. Its cap is 50 without pagination; result_capped means the local result is incomplete. Distinguish a failed geocode (place_not_found) from a successful search with no returned branches. Canadian operational data should use the Canadian host when requested.
Other brands can use stored Google Maps business search/facets/nearby, followed by selected live search/place verification. A name-text match is not a verified chain affiliation; check canonical brand/site, address, and location evidence. Google Maps category/country values differ from Starbucks taxonomy/codes.
McDonald's coordinate locator is a separate source with its own accepted markets, cap, and radius in miles; dataset nearby radii are meters. Convert explicit geographic scopes rather than copying identical numeric radius values. A capped live locator alone cannot establish a whole-country store count.
Keep a boundary/query ledger, requested/returned pages, cap flags, search result counts, and crawl times. The stored indexes are not automatically an official census of every branch or a proof that a competing chain is absent.
scripts/crawlora.sh /datasets/starbucks-stores/facets facet=country
scripts/crawlora.sh /datasets/starbucks-stores/search country=US city=Seattle page=1 page_size=100
scripts/crawlora.sh /starbucks/stores place="Seattle, WA" market=us
scripts/crawlora.sh /datasets/google-map-businesses/search q="Starbucks" city=Seattle page=1 page_size=10

Reconcile locations and report coverage

Deduplicate the same brand/store ID first; cross-source matches need address and coordinate corroboration. Similar names are not sufficient, and two units/kiosks inside a complex can be distinct branches. Keep relocated, permanently-closed, licensed, and company-operated locations explicit when fields support the labels. An empty schedule/amenity field can mean not published for that market, not closed or unsupported. Mobile-order readiness differs from static capability codes.

Use unique matched branches for area counts. For overlapping radius queries, count each branch once and keep the union boundary explicit; dividing by the sum of overlapping circle areas produces a misleading density. Area density requires a defined area, and per-capita measures require a separately sourced population with matching geography/date. Counts alone are not market share.

Return a branch evidence ledger, deduplicated area/brand matrix, source coverage, observed amenity/ownership breakdowns, and unknown or incomplete areas. A missing branch in a capped/partial sample is an observed coverage gap, not proof of closure or unmet demand. Historical expansion/contraction claims require comparable saved observations and verified changes. Do not contact stores, submit leads, or choose an investment/site solely from this footprint.

开发者人才发现
Xbox 游戏研究