Shufersal
This Shufersal MCP server provides automated shopping capabilities for the Israeli supermarket chain Shufersal through browser automation using Puppeteer, developed by Mati Horowitz. The implementation enables AI assistants to create shopping lists from recipe links, search the Shufersal product catalog with personalized results for logged-in users, and add items directly to the shopping cart through a headless browser that maintains user session data. Built with TypeScript and the MCP SDK, it offers specialized tools for reading webpage content and converting it to shopping lists, making it valuable for meal planning workflows where users want to automatically populate their Shufersal cart from recipe ingredients found online.
I/O
heuristic · beta该 skill 无 npm 包，暂用描述启发式推断 I/O；接入后可实测。
综合分
效果测评卡
进行中 · 待运行真实 I/O 已解决「输入输出是什么」；下一步接 agent 真跑，量化「效果好不好」。
Profitlee calculates ecommerce profit margins for Amazon FBA, FBM, and TikTok Shop, breaking down fees and providing scenario planning. The base profit calculation works without an API token; advanced scenario tools require a Profitlee Pro subscription.
This AIWerk integration connects AI agents to the Shopify Admin GraphQL API for managing products, orders, customers, and inventory with full read and write access. Configuration requires a Shopify store domain and Admin API app credentials. The server is installed locally via npm.
Hookwarden is a local MCP server that exposes a scan_handler tool for auditing webhook handler code. It checks signature-verification logic for correctness across 20+ providers including Stripe, GitHub, Shopify, and Twilio, returning a three-state verdict without sending any data off-machine. The underlying engine is pure-functional with no I/O or network calls, making results fully deterministic and reproducible across CI, CLI, and MCP environments.