Eko

Stale
GitHub TypeScript MIT

Description

Build production-ready agentic workflows with natural language, supporting browser automation, computer use, and RAG workflows

Key Features

  • Production-ready JavaScript framework for building reliable agents with natural language
  • Multi-agent support with task parallelization and dependency-aware execution
  • Native MCP protocol integration for seamless tool connectivity
  • Human-in-the-loop controls with pause, resume, and interrupt capabilities
  • Cross-platform runtime: browser extensions, Node.js, and web applications
  • Dynamic LLM switching with support for OpenAI, Anthropic, Google, and compatible providers

Use Cases

💡 Browser automation and intelligent web scraping workflows
💡 System file and process management via natural language commands
💡 Multi-step task orchestration across desktop and browser environments
💡 Data processing, organization, and reporting automation
💡 RAG workflow automation with document retrieval and synthesis

Tags

workflow browser-automation computer-use natural-language agentic typescript

Categories

🔄 Workflow ⚡ Agent Tools

Quick Start

Install: pnpm install @eko-ai/eko. Configure LLM providers (Anthropic, OpenAI, Google), create agents (BrowserAgent, FileAgent), and run: const eko = new Eko({ llms, agents }); await eko.run("Search for Musk news, save to desktop"). See eko.fellou.ai/docs for full guide.
Visit GitHub Visit Website

Related Projects