Keystroke 是一个开源 TypeScript 框架和托管平台,用于从编码智能体(如 Claude Code、Cursor、Codex)直接构建 AI 智能体和工作流自动化。

Stars

46

7 天增长

暂无数据

Fork 数

9

开放 Issue

0

开源协议

NOASSERTION

最近更新

2026-07-24

AI 仓库情报摘要
FR-AI / ANALYSIS

为什么值得关注

它将自动化视为真实的、类型化的、经过测试和 Git 版本管理的 TypeScript 代码,而非 JSON 工作流文件,连接了 AI 智能体开发与传统软件工程实践。

适合谁使用

  • 使用 AI 编码智能体(Claude Code、Cursor、Codex)的开发者
  • 构建生产级 AI 自动化和工作流的团队
  • 偏好代码驱动、可测试的自动化而非低代码画布的开发者
  • 寻求自托管或云管理的智能体编排的平台工程师

典型使用场景

  • 创建每日简报智能体,扫描日历、电子邮件和问题后发布到 Slack
  • 构建数据分析师智能体,通过 Slack 连接数据仓库回答临时查询
  • 定义基于 Cron 触发的持久化工作流,支持步骤重试和幂等执行

项目优势

  • 自动化是纯 TypeScript(智能体、动作、工作流),支持完整的 IDE 功能、类型检查和可测试性
  • 持久化步骤:成功的步骤会记录并在重试时回放,而非重新执行
  • 1000+ 集成定义,支持任何 HTTP API 或 MCP 服务器
  • 双部署模式:开源独立版本用于本地/私有使用,托管云版本用于团队协作

使用前须知

  • 采用 Elastic License 2.0 许可,禁止将核心功能作为托管服务提供给第三方
  • 需要 Node.js 20+,限制了在较旧环境中的使用
  • 依赖编码智能体(Claude Code、Cursor、Codex)进行初始设置和提示驱动开发

README 快速开始

AI automation, built for your coding agent

Build specialized agents and workflow automations with Claude Code, Cursor, Codex, or any coding agent. Everything is real TypeScript in your repo.

Get started · Documentation · Use cases


Keystroke is what n8n would look like if it were built for coding agents. Describe the agent or automation you want, and your coding agent builds it as typed, tested, git-versioned TypeScript. Deploy it to a shared platform where your team can run it, connect credentials, and inspect every run.

No JSON workflow blobs. No canvas as the source of truth. Your automation is code your agent can grep, test, review, and fix.

Start with your coding agent

Paste this into Claude Code, Cursor, Codex, or another agentic IDE:

Read https://keystroke.ai/start.md then help me create my first automation.

Your agent installs the CLI, connects your account, scaffolds a project, and deploys the first version. Then describe what you want to build:

Each weekday morning, scan my Google Calendar, Gmail inbox, and Linear issues,
then send me a short briefing in Slack.
Build a data analyst agent that answers ad hoc business questions by querying
our warehouse. Connect it to Slack so the team can ask questions in #data.

Browse more prompts →

Or set it up yourself

Prerequisite: Node.js 20 or later.

npm install -g @keystrokehq/cli
keystroke auth login
keystroke init my-app --yes
cd my-app
keystroke projects list
keystroke projects link --project 
keystroke deploy

keystroke deploy lints, typechecks, builds, and uploads your project. No separate build steps are required.

Everything is TypeScript

Agents, workflows, actions, and triggers live under src/ in your project.

my-app/
├── src/
│   ├── actions/
│   ├── agents/
│   ├── triggers/
│   └── workflows/
├── AGENTS.md
├── keystroke.config.ts
├── package.json
└── tsconfig.json

Define an agent:

import { defineAgent } from "@keystrokehq/keystroke/agent";

export default defineAgent({
  slug: "researcher",
  name: "Researcher",
  description: "Researches a topic and returns a sourced brief.",
  systemPrompt: "Research the request, verify important claims, and cite sources.",
  model: "anthropic/claude-sonnet-4.6",
});

Create a reusable action

相关仓库与替代方案

根据分类、Topic 和编程语言匹配的相似项目。

TanStack
精选
TanStack GitHub avatar

router

TanStack Router is a type-safe, data-driven React router with built-in caching, prefetching, and nested layouts, while TanStack Start extends it into a full-stack SSR framework.

Web 开发前端框架
14,861
vercel-labs
精选
vercel-labs GitHub avatar

scriptc

scriptc compiles ordinary TypeScript into small, fast native executables without needing Node.js, V8, or any JavaScript runtime in the binary.

开发者工具代码质量与构建
1,985
Jakubantalik
精选
Jakubantalik GitHub avatar

thinking-orbs

A React component library that renders six hand-tuned animated thought orb loading indicators on a plain 2D canvas, with two purpose-tuned sizes and automatic theme detection for AI and agent UIs.

AI 与机器学习AI 智能体
1,191