← 返回 Skills 市场

Housesigma Collector

Name: Housesigma Collector
Author: sonim1

作者 sonim1 · GitHub ↗ · v1.0.0

cross-platform ⚠ suspicious

356

总下载

当前安装

版本数

在 OpenClaw 中安装

/install housesigma-collector

功能描述

HouseSigma 매물 데이터를 자동 및 수동으로 수집해 Hauscout SQLite DB에 저장하고 크론잡으로 주기적 업데이트를 지원합니다.

使用说明 (SKILL.md)

HouseSigma Collector Skill

HouseSigma에서 매물 데이터를 수집하여 Hauscout SQLite DB에 저장하는 skill.

프로젝트 경로

Hauscout: /Users/kendrick/projects/hauscout
DB: /Users/kendrick/projects/hauscout/data/hauscout.db
스크립트: /Users/kendrick/projects/hauscout/scripts/collect.ts

사용법

자동 수집 (검색 프로필 기반)

cd /Users/kendrick/projects/hauscout && npx tsx scripts/collect.ts

특정 매물 수집

cd /Users/kendrick/projects/hauscout && npx tsx scripts/collect.ts --url "\x3Chousesigma_url>"

특정 프로필만 수집

cd /Users/kendrick/projects/hauscout && npx tsx scripts/collect.ts --profile \x3Cid>

브라우저 창 보면서 수집 (디버깅)

cd /Users/kendrick/projects/hauscout && npx tsx scripts/collect.ts --headed

수동 수집 (브라우저 직접 사용)

Playwright 스크립트 대신 Clawdbot 브라우저로 직접 수집할 때:

브라우저로 HouseSigma 매물 상세 페이지 열기
DOM 스냅샷에서 데이터 추출
SQLite에 직접 INSERT

DOM 데이터 매핑

HouseSigma 상세 페이지의 구조:

주소/상태: \x3Ch1> 태그 (Unit X - Street - Municipality - Community)
가격: \x3Cem> 태그의 $ X,XXX 패턴
Key Facts: \x3Cdt>/\x3Cdd> 쌍 (Tax, Property Type, Maintenance, etc.)
Details: 같은 \x3Cdt>/\x3Cdd> 패턴 (Bedrooms, Bathrooms, etc.)
방 정보: "Metres" 섹션의 텍스트 패턴
Estimates: SigmaEstimate, Estimated Rent, Rental Yield
학교: Catchment Schools 섹션
인기도: "Popularity : XX/100" 텍스트
커뮤니티 통계: Community Statistics 섹션

검색 프로필 관리

프로필 추가:

cd /Users/kendrick/projects/hauscout
sqlite3 data/hauscout.db "INSERT INTO search_profiles (name, center_lat, center_lng, radius_km, property_types, price_min, price_max, beds_min, beds_max, baths_min, baths_max, is_active) VALUES ('이름', lat, lng, radius, '[\"Condo Apartment\"]', 0, 800000, 2, 3, 1, 2, 1);"

현재 프로필 확인:

sqlite3 data/hauscout.db "SELECT * FROM search_profiles;"

데이터 수집 후

수집 후 대시보드에 반영하려면:

cd /Users/kendrick/projects/hauscout
git add data/hauscout.db
git commit -m "data: daily collection $(date +%Y-%m-%d)"
git push

Vercel 배포 시 자동으로 최신 데이터가 반영됩니다.

크론잡

Clawdbot cron으로 매일 오전 9시에 자동 수집:

스크립트 실행 → DB 업데이트 → git commit & push

주의사항

HouseSigma rate limiting 방지를 위해 요청 간 2-4초 간격 유지
headless 모드에서 차단될 수 있음 → --headed 옵션으로 확인
검색 프로필의 결과가 많으면 시간이 오래 걸림 (매물당 ~5초)

安全使用建议

This skill appears to do what it says (collect HouseSigma listings into a local Hauscout SQLite DB), but review several things before installing or running it: 1) Inspect scripts/collect.ts and package.json in the referenced project to ensure the code is safe and dependencies are trusted — npx tsx will fetch/execute code. 2) Do not commit and push the SQLite DB unless you are sure it contains no sensitive data; consider adding data/hauscout.db to .gitignore or using a scrubbed export for publishing. 3) The SKILL.md uses absolute paths tied to a specific user; update paths to match your environment. 4) Be aware that running the script will use your git credentials and network access; if you need to avoid accidental remote pushes, run in an isolated environment or remove the git push step. 5) Check HouseSigma's terms of service and robots/rate-limiting policies before scraping. If you can provide the actual scripts/collect.ts (and package.json) or a vetted install spec, I can give a higher-confidence assessment.

功能分析

Type: OpenClaw Skill Name: housesigma-collector Version: 1.0.0 The skill bundle automates real estate data collection from HouseSigma using shell commands, SQLite manipulation, and Git operations. While the instructions in SKILL.md are consistent with the stated purpose of updating a Vercel-hosted dashboard, the skill employs high-risk capabilities including automated 'git push' and execution of local TypeScript scripts via 'npx tsx'. The presence of hardcoded absolute paths (e.g., /Users/kendrick/projects/hauscout) suggests it is tailored for a specific environment, but the broad shell access and automated repository synchronization are risky behaviors that warrant caution.

能力评估

ℹ Purpose & Capability

The name/description (collect HouseSigma data into Hauscout SQLite DB) matches the runtime instructions: cd into a local project, run scripts/collect.ts, and insert into hauscout.db. The use of sqlite3 and browser DOM extraction is coherent with the purpose. However the SKILL.md hard-codes absolute user-local paths (/Users/kendrick/...), which makes it tied to a specific machine and may not apply to other users.

⚠ Instruction Scope

Instructions tell the agent to read and update a local SQLite DB, run a TypeScript script via npx tsx, and then git add/commit/push the DB so it becomes available via Vercel. Automating git push of a database can cause unintended data exfiltration; the manual-collection instructions also direct DOM scraping and direct INSERTs into the DB. The steps are within the stated purpose, but the DB commit/push and lack of guidance about what is safe to push are risky.

ℹ Install Mechanism

There is no explicit install spec (instruction-only). The runtime commands rely on npx tsx and sqlite3/git being present. Using npx will fetch and run packages from npm at runtime which can execute arbitrary code — this is expected for running a local TypeScript script but is a noteworthy risk if you haven't audited scripts/collect.ts and package dependencies.

ℹ Credentials

The skill declares no required environment variables or credentials, which is consistent. However, the workflow implicitly uses system git credentials (for git push), network access to npm (npx), and possibly browser automation credentials/session state. Requiring a commit & push of the DB is disproportionate to mere collection unless the user explicitly intends to publish that DB; it may leak sensitive data if present.

✓ Persistence & Privilege

always is false and the skill is instruction-only with no install; it does not request permanent presence or modify other skills. Autonomous invocation is allowed (platform default) but not a special privilege here. The main privilege is being asked to operate on local files and run commands when invoked.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install housesigma-collector
安装完成后，直接呼叫该 Skill 的名称或使用 /housesigma-collector 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

Initial release of the HouseSigma Collector Skill. - Collects HouseSigma listing data and saves to Hauscout SQLite DB. - Supports automated, profile-based, and targeted collection via command-line options. - Provides manual collection workflow using browser and DOM data mapping for custom data extraction. - Includes instructions for managing search profiles and integrating collected data with the dashboard. - Supports daily automated collection via cron, with built-in precautions for rate limiting and headless browsing. - Detailed usage and maintenance guides are provided in the documentation.

元数据

Slug housesigma-collector

版本 1.0.0

许可证 —

累计安装 1

当前安装数 1

历史版本数 1

常见问题