gr-scraper-mcp
Tracks live board game inventory on thegamerules.com by scraping the catalogue, diffing against the last snapshot, and reporting additions/removals. Integrates with a Telegram bot for commands and notifications.
README
gr-scraper-mcp
MCP server that tracks live board game inventory on thegamerules.com. On each run it scrapes the full catalogue, diffs it against the last stored snapshot, and reports what was added and what was removed. Deployed on Render, backed by Upstash Redis, with a Telegram bot interface.
How it works
- Puppeteer launches a headless Chromium browser and visits the product listing pages
- Images, fonts, stylesheets, and tracking scripts are blocked to speed up rendering
- The scraper waits for
.nameelements to appear, then collects all game titles - The live list is compared against the snapshot stored in Redis
- Added and removed games are returned, and the snapshot is updated
MCP tool
| Tool | Description |
|---|---|
check_game_changes |
Scrape live games, diff against stored snapshot, update storage. Returns added, removed, total, stored_before. |
Telegram bot
| Command | Action |
|---|---|
/check |
Trigger a scrape and reply with what changed |
/status |
Show stored game count without scraping |
| Any other message | Treated as a /check request |
If TELEGRAM_CHAT_ID is set, the bot also pushes a notification whenever the MCP tool is called from Claude.
Environment variables
| Variable | Required | Description |
|---|---|---|
UPSTASH_REDIS_REST_URL |
✅ | Upstash Redis REST endpoint |
UPSTASH_REDIS_REST_TOKEN |
✅ | Upstash Redis REST token |
TELEGRAM_BOT_TOKEN |
✅ | Token from @BotFather |
TELEGRAM_CHAT_ID |
Optional | Chat ID for push notifications |
Setup
See DEPLOY.md for full step-by-step deployment instructions.
Local development
cp .env.example .env # fill in your credentials
npm install
npm start # requires Chromium at /usr/bin/chromium
# or set PUPPETEER_EXECUTABLE_PATH
Stack
- Runtime: Node.js 20
- MCP:
@modelcontextprotocol/sdkover HTTP (Streamable HTTP transport) - Browser: Puppeteer Core + system Chromium
- Storage: Upstash Redis
- Bot:
node-telegram-bot-api(webhook mode) - Deployment: Render (Docker, web service)
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。