// the find
NaiboWang/EasySpider
A visual no-code/code-free web crawler/spider易采集：一个可视化浏览器自动化测试/数据采集/网页爬虫软件，可以无代码图形化的设计和执行爬虫任务。别名：ServiceWrapper面向Web应用的智能化服务封装系统。
EasySpider is an Electron-based, no-code visual web scraper: you click elements on a live page, it infers repeated patterns (product cards, titles, pagination), and builds a flowchart-style task you can run from the GUI or headless via CLI. It's aimed at non-programmers and ops/data people who need structured scraping without writing Scrapy/Playwright code, and it has real academic backing (WWW 2023 paper).
The element-selection UX is more than a gimmick — auto-detecting sibling blocks, selecting sub-elements, and 'loop-click into detail page' cover the actual annoying parts of scraping listing+detail sites, not just static tables. Tasks defined in the GUI can be executed headlessly from the command line, so once a non-technical user builds a task, an engineer can cron it or embed it in a pipeline. It also bakes in things people usually bolt on later: custom JS injection mid-flow, IP/proxy rotation, and OCR hooks for text captchas.
It's AGPL-3.0, which the README itself spells out at length: if you modify it and run it as a service (SaaS/API), you're obligated to publish your full modified source to users — a real constraint for anyone wanting to embed this commercially rather than just run the desktop app. There's no clean programmatic task API; tasks are opaque JSON blobs, and the repo tree shows hundreds of them (0.json through 252.json, duplicated across Linux/macOS folders) committed directly into .temp_to_pub, which is repo clutter rather than sample data done well. Documentation is primarily a Chinese GitHub Wiki and Bilibili videos, with English as a secondary pass, so debugging an edge case in English is harder than the star count suggests. The README also spends more space on proxy/captcha-solving affiliate placements (Bright Data, CapSolver, Webshare, Thordata) than on architecture or contribution docs.