#140140th of 142 in Agent frameworks and SDKs
eko
JavaScript framework for natural-language agent workflows in browsers and Node.js
- Stars
- 5k
- License
- MIT
- Last commit
- Mar 2026
- Last release
- Dec 2025
- Language
- TypeScript
4,850 on 22 Apr 20264,964 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Eko is a TypeScript/JavaScript framework that turns a natural-language task into a multi-step agent workflow. It runs in browser extensions, web apps and Node.js, with built-in browser and file agents, MCP tool support, and configurable LLM providers such as Anthropic, Google, OpenAI and OpenAI-compatible endpoints. Version 3.0 added dependency-aware parallel execution, plus pause, resume and interrupt controls.
Who it is for: JavaScript developers building browser or Node.js automation agents
Strengths
- Runs in browser extensions, web apps and Node.js through one API
- Native MCP support and custom agents or tools
- Parallel agent execution, pause/resume and task_snapshot workflow recovery
- Works with Anthropic, Google, OpenAI and OpenAI-compatible endpoints like Qwen
Weaknesses
- Browser use needs a backend proxy to keep API keys out of frontend code
- Observable chain and native A2A are listed as not yet implemented
- Upgrading from 3.x requires regenerating saved workflows to the new schema
- Comparison table in the README is self-reported and unsourced
What it needs
- no GPU
- Needs Node.js, pnpm, Playwright (Node.js example), LLM provider API key
- Models: Anthropic, Google Gemini, OpenAI, OpenAI-compatible (Qwen, Doubao)
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in Agent frameworks and SDKs
See all 142| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 96 out of 100 |
| 2 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 92 out of 100 |
| 3 | mastraTypeScript framework for building AI agents, workflows and MCP servers | 92 out of 100 |
| #4 | browser-usePython library that lets an LLM agent control a web browser | 90 out of 100 |
| #5 | copilotkitReact and Angular SDK for adding agent chat and generative UI to apps | 90 out of 100 |