能力标签

⚙️

Agent工作流

MoMoA研究员

基于 TypeScript · 无代码搭建完整 AI 自动化流程

英文名：MoMoA-Researcher

⭐ 60 Stars 🍴 15 Forks 💻 TypeScript 📄 Apache-2.0 🏷 AI 8.0分

8.0AI 综合评分

AI工作流TypeScript

⬇ 下载源码 ZIP ⚙️ 配置说明

✦ AI Skill Hub 推荐

经 AI Skill Hub 精选评估，MoMoA研究员获评「强烈推荐」。这款Agent工作流在功能完整性、社区活跃度和易用性方面表现出色，AI 评分 8.0 分，适合有一定技术背景的用户使用。

📚 深度解析

MoMoA研究员是一套完整的 AI Agent 自动化工作流方案。随着 AI 能力的不断提升，基于 Agent 的自动化工作流正在成为提升个人和团队效率的核心方式。区别于传统的 RPA 自动化（模拟鼠标键盘操作），AI Agent 工作流通过理解任务意图、动态规划执行路径，能够处理更复杂的非结构化任务。

MoMoA研究员工作流的设计遵循"最小配置，最大复用"原则：核心逻辑已经封装好，用户只需配置自己的 API Key 和业务参数即可快速上手。工作流内置错误处理和重试机制，在网络波动或 API 限速等情况下仍能稳定运行，适合作为生产环境的自动化基础设施。

在实际部署时，建议先在测试环境中运行 3-5 次，验证各个环节的输出结果符合预期，再部署到生产环境。AI Skill Hub 评分 8.0 分，是同类 Agent 工作流中的精选推荐。

📋 工具概览

MoMoA研究员是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排，将复杂的多步骤任务拆解为清晰的自动化流程，实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成，适合构建数据处理管线、业务自动化和 AI 辅助决策系统。

GitHub Stars

⭐ 60

开发语言

TypeScript

支持平台

Windows / macOS / Linux

维护状态

轻量级项目，按需更新

开源协议

Apache-2.0

AI 综合评分

8.0 分

工具类型

Agent工作流

Forks

📖 中文文档

以下内容由 AI Skill Hub 根据项目信息自动整理，如需查看完整原始文档请访问底部「原始来源」。

📌 核心特色

可视化 Agent 工作流编排，无需编写复杂代码
支持多步骤自动化任务链，实现全流程无人值守
与外部 API、数据库和第三方服务无缝集成
内置错误处理与自动重试机制，保障稳定运行
提供可复用的自动化模板，快速在同类场景部署

🎯 主要使用场景

自动化日常重复性工作，将精力集中于创造性任务
构建数据采集 → 处理 → 输出的完整自动化管线
实现跨平台、跨系统的数据流转和业务协同

以下安装命令基于项目开发语言和类型自动生成，实际以官方 README 为准。

安装命令

# 方式一：npm 全局安装
npm install -g momoa-researcher

# 方式二：npx 直接运行（无需安装）
npx momoa-researcher --help

# 方式三：项目依赖安装
npm install momoa-researcher

# 方式四：从源码运行
git clone https://github.com/retomeier/MoMoA-Researcher
cd MoMoA-Researcher
npm install
npm start

📋 安装步骤说明

访问 GitHub 仓库获取工作流文件
在对应平台（Dify / Flowise / Make 等）中找到「导入工作流」功能
上传工作流文件
按照提示配置必要的环境变量和 API Key
运行测试确认流程正常后投入使用

以下用法示例由 AI Skill Hub 整理，涵盖最常见的使用场景。

常用命令 / 代码示例

# 命令行使用
momoa-researcher --help

# 基本用法
momoa-researcher [options] <input>

# Node.js 代码中使用
const momoa_researcher = require('momoa-researcher');

const result = await momoa_researcher.run(options);
console.log(result);

以下配置示例基于典型使用场景生成，具体参数请参照官方文档调整。

配置示例

# momoa-researcher 配置说明
# 查看配置选项
momoa-researcher --config-example > config.yml

# 常见配置项
# output_dir: ./output
# log_level: info
# workers: 4

# 环境变量（覆盖配置文件）
export MOMOA_RESEARCHER_CONFIG="/path/to/config.yml"

📑 README 深度解析真实文档完整度 32/100 查看 GitHub 原文 →

以下内容由系统直接从 GitHub README 解析整理，保留代码块、表格与列表结构。

Mixture of Mixture of Agents - Researcher

Coordinate multiple AI agents to iteratively investigate complex research questions by proposing hypotheses, executing local Python or Rust experiments, evaluating outcomes, and generating academic reports.

----

MoMoA (Mixture of Mixture of Agents) Researcher is a full-stack application. The primary workflow involves defining a research objective in the React frontend, which pushes task definitions to a Firebase Realtime Database. A Node.js server listens for these queues, cloning required repositories into memory and spinning up specialized agent roles to run experiments.

It is an extension / adaptation of the Mixture of Mixture of Agents SDLC Agent.

The DashboardPage.tsx shows how a research session is started from the User Interface. The ResearchProjectPage.tsx shows how session updates are displayed to the user, and implements the continuous research loop.

----

This system builds on the Experts, Work Phases, and Tools of the MoMoA Agent. It adds a "Senior Researcher", "Research Room", and Research Tools to conduct experiments using the scientific method.

About this Project

Project Home Page: https://labs.google/code/experiments/momoa-researcher

Code Home: https://github.com/retomeier/MoMoA-Researcher

Maintained by: Reto Meier

UI & Orchestration Features

The React frontend enables project and task initiation, visualizes the agent's work, and facilitates human oversight.

Research Projects: A hierarchical workspace where an overarching goal (e.g., "Find the longest Collatz sequence") contains multiple individual research sessions/tasks.
Interrogation Chat: A chat interface that persists the full context of previous tasks, modified files, and session logs, allowing you to question the agent about its methodologies during or after a run.
Self-Evaluation & Continuous Loop: After a task, the agent grades its own performance and proposes the next logical experiments. Toggling "Auto-run top suggestion" will execute these sequentially without human intervention.

Setup & Configuration

This project is built using a Node.js backend (tsx, express) and a React/Vite frontend. It relies heavily on Firebase Realtime Database for state synchronization between the client and the orchestrator.

1. Clone and Install Dependencies The project uses concurrently to run both web and server environments.

npm install

2. Configure Firebase

This project uses Firebase to store details on each project and session. To use the UI, you will need to create your own Firebase project and provide your own client and server-side configuration details.

2.1. Create a Firebase Project

Go to the Firebase Console.
Click Add project and follow the on-screen instructions.

2.2. Register a Web App

In your new Firebase project dashboard, click the Web icon (</>) to add a new web app.
Give your app a nickname and click Register app.

2.3. Get Your Client Configuration

After registering, Firebase will provide you with a Firebase Config object containing your API keys and identifiers. This should be placed in the web/src/firebase-config.ts file:

2.4. Generate the Service Account Key (Admin SDK) Now create server-side administrative access: In the Firebase Console, click the gear icon next to Project Overview in the top left and select Project settings. Navigate to the Service accounts tab. Click the Generate new private key button at the bottom, then click Generate key to confirm. A JSON file containing your service account credentials will securely download to your machine.

2.5. Add the Service Account File Rename the downloaded JSON file to .firebase-service-account.json. Move this file into the root directory of your local project repository. * Security Warning: Never commit this file to version control. It grants full administrative control over your Firebase project database and auth.

2.6. Link the Firebase CLI (Optional but required for deploying/functions) If you need to deploy Firebase Functions, Security Rules, or Hosting, you must link your local environment to your Firebase project using the Firebase CLI. If you don't have the CLI installed, install it by running: npm install -g firebase-tools Log in to the Firebase CLI:

firebase login

Link this local directory to your Firebase project:
firebase use --add
When prompted, select the Firebase project you created in Step 1 and type default as the alias. This will automatically generate a .firebaserc file for you.

3. Run the application Start the frontend and backend simultaneously: