Happy slice of mcp pizzaMCP.pizza

ediora-studio

MCP.Pizza Chef: phoenix-gh

Ediora Studio creates a workspace where you can gather information from various sources, select valuable content, and use AI to help write and publish articles or posts. It works with ChatGPT and other AI providers, letting you manage your content creation all in one place. Setup requires some developer knowledge, including Docker and environment setup, but once running, it streamlines your AI writing workflow.

Other
Web/Research
Writing

Use This MCP client To

Collect information from web feeds like RSS and YouTube Filter and select valuable content automatically Create AI-generated articles and posts Manage and publish AI-assisted content Configure multiple AI providers for content creation Track and schedule content tasks Use a workspace that integrates AI writing and publishing

README

Ediora · 述策

AI 内容工作台：集信息采集、价值甄选、AI 创作与发布辅助于一体。

技术栈

  • 后端：FastAPI + PostgreSQL（SQLAlchemy async）+ APScheduler，Python 3.11（conda env: wems）
  • 前端：Next.js App Router (16.x) + Tailwind CSS + shadcn/ui，包管理器 pnpm
  • AI 层：多 Provider（OpenAI / Anthropic / DeepSeek / 通义千问 / Moonshot / GLM / MiniMax / MiMo / 自定义）
  • X 采集：feedgrab（Cookie 认证）
  • 任务执行：PostgreSQL 持久化任务 + Redis 队列 + Next.js AI SDK worker

目录结构

Ediora/
├── backend/
│   ├── main.py                 应用入口（瘦版本，94 行）
│   ├── scheduler.py            APScheduler 任务注册 + 节流
│   ├── models.py               SQLAlchemy ORM
│   ├── database.py             数据库连接（默认 Postgres）
│   ├── config.py               app_settings 键值配置（60s 缓存）
│   ├── collector.py            RSS/YouTube/Reddit 通用采集
│   ├── llm.py                  多 Provider LLM 调用
│   ├── x_response_service.py   X 即时响应策略、幂等任务与补偿
│   ├── x_response_links.py     外链安全核验
│   ├── telegram_notifier.py    Telegram Bot API 推送
│   ├── *_collector.py          站点专用采集器（juejin/kr/v2ex/wechat/x/youtube/...）
│   ├── mcp_server.py           MCP streamable HTTP server
│   └── routers/                FastAPI 路由（统一 prefix=/<module>，main 处加 /api）
│
├── web/
│   ├── app/                    页面（drafts/topics/quotes/github/papers/...）
│   ├── components/
│   │   ├── features/           业务组件（Sidebar/ArticleReader/...）
│   │   └── ui/                 shadcn/ui
│   └── lib/
│       ├── api/                各模块 API 客户端
│       ├── format.ts           共享格式化 helper（fmtRelTime/fmtNum/fmtCount/fmtShortDate/fmtFullDate）
│       ├── use-infinite-scroll.ts 共享无限滚动 hook
│       └── use-media-query.ts  媒体查询 hook
│
└── chrome-plug/                Chrome 插件

启动方式

宿主机开发推荐从项目根目录使用统一脚本。先确保 Docker、conda 环境 wems、pnpm 和 redis-server（仅在本机没有可用 Redis 时需要）可从 当前 PATH 找到。首次使用时从示例创建根环境文件，并在其中配置 API 与 内容任务 worker 共用的 32 字符以上 token：

ediora-studio FAQ

Can I use ediora-studio with ChatGPT?
Yes — it supports ChatGPT as a client app to enhance your AI content creation.
Can I use ediora-studio to gather and create AI content?
Yes — it helps collect information and assist AI-powered writing and publishing.
Do I need an API key or account to use ediora-studio?
You need to configure API keys for AI providers like OpenAI, Anthropic, or others within the app settings.
How hard is it to set up ediora-studio?
Setup requires developer skills, including Docker and environment configuration.
Does ediora-studio work with other AI providers besides OpenAI?
Yes — it supports multiple providers including Anthropic, DeepSeek, and others.
Can I install ediora-studio easily on my computer?
Installation involves developer steps like running Docker and configuring environment variables.