经 AI Skill Hub 精选评估，原子工作流 获评「推荐使用」。这款MCP工具在功能完整性、社区活跃度和易用性方面表现出色，AI 评分 7.5 分，适合有一定技术背景的用户使用。
📚 深度解析
通过安装 原子工作流，你的 AI 助手将获得额外的工具调用能力，可以用自然语言直接操控该工具的功能，无需学习复杂的命令行语法。MCP 工具的核心价值在于"一次配置，永久增强"——配置完成后，每次与 AI 对话时都可以无缝调用这些工具。
在技术实现上，MCP 工具通过标准的 JSON-RPC 协议与 AI 客户端通信，工具的功能以"工具列表"的形式暴露给 AI 模型，AI 可以按需调用。原子工作流 提供了结构化的工具调用接口，使 AI 模型能够精确地理解和使用每个功能点，显著降低 AI 在工具使用上的错误率。
与传统的 API 集成相比，MCP 工具的优势在于无需编写代码——用户只需在配置文件中添加几行 JSON，即可让 AI 获得全新能力。AI Skill Hub 将 原子工作流 评为 AI 评分 7.5 分，属于同类工具中的优质选择。
📋 工具概览
原子工作流 是一款遵循 MCP（Model Context Protocol）标准协议的 AI 工具扩展。通过 MCP 协议，它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务，实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用，都可以通过自然语言在 AI 对话中直接触发，极大提升生产效率。
📖 中文文档
原子工作流 是一款遵循 MCP（Model Context Protocol）标准协议的 AI 工具扩展。通过 MCP 协议，它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务，实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用，都可以通过自然语言在 AI 对话中直接触发，极大提升生产效率。
- 通过标准 MCP 协议与 Claude、Cursor 等主流 AI 客户端深度集成
- 提供结构化工具调用接口，显著降低 AI 集成复杂度
- 支持 Claude Desktop 和 Claude Code 无缝接入，开箱即用
- 可与其他 MCP 工具组合叠加，构建完整 AI 工作站
- 轻量无侵入设计，不影响现有系统架构
- 在 Claude Desktop 对话中直接调用本地工具，实现 AI 与系统的深度联动
- 通过自然语言驱动复杂的多步骤自动化任务，代替繁琐手动操作
- 将多个 MCP 工具组合使用，构建个人专属 AI 工作站
# 方式一：通过 Claude Code CLI 一键安装
claude skill install https://github.com/bastani-inc/atomic
# 方式二：手动配置 claude_desktop_config.json
{
"mcpServers": {
"-----": {
"command": "npx",
"args": ["-y", "atomic"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
- 确认已安装 Node.js（v18 或以上版本）
- 打开 Claude Desktop 或 Claude Code 的 MCP 配置文件
- 按「交给 Agent 安装 → Claude Desktop」标签中的 JSON 配置填入 mcpServers 字段
- 保存配置文件并重启 Claude 客户端
- 重启后，在对话中即可使用本工具
# 安装后在 Claude 对话中直接使用 # 示例： 用户: 请帮我用 原子工作流 执行以下任务... Claude: [自动调用 原子工作流 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入："列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"_____": {
"command": "npx",
"args": ["-y", "atomic"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
简介
Atomic - Dynamic Workflows for Software Engineering
<p align="center"><img width="800" height="450" alt="atomic-promo" src="./assets/atomic-promo.gif" /></p>
<p align="center"> <b>Atomic is the workflow layer for coding agents, giving developers a programmable control plane for complex engineering work.</b><br> <i>An open-source, model-agnostic take on dynamic workflows for software engineering — with Pi extensions, custom models, MCP, sub-agents, artifacts, and review gates.</i> </p>
<p align="center"> <a href="#get-started"><b>Get started →</b></a> · <a href="#spec-driven-development">Spec-driven development</a> · <a href="#when-should-i-use-atomic">When to use Atomic</a> · <a href="#faq">FAQ</a> · <a href="https://docs.bastani.ai/">Docs</a> </p>
<p align="center"> <a href="https://docs.bastani.ai/"><img src="https://img.shields.io/badge/docs-atomic-blue" alt="Docs"></a> <a href="https://discord.gg/9CvdXUGXR4"><img src="https://img.shields.io/badge/join%20community-discord-5865F2?logo=discord&logoColor=white" alt="Discord"></a> <a href="https://deepwiki.com/bastani-inc/atomic"><img src="https://deepwiki.com/badge.svg" alt="Ask DeepWiki"></a> <a href="./package.json"><img src="https://img.shields.io/badge/TypeScript-6.x-3178C6?logo=typescript&logoColor=white" alt="TypeScript"></a> <a href="./package.json"><img src="https://img.shields.io/badge/Bun-Runtime-f9f1e1?logo=bun&logoColor=black" alt="Bun"></a> <a href="./LICENSE"><img src="https://img.shields.io/badge/License-MIT-yellow.svg" alt="License: MIT"></a> </p>
---
Prerequisites
- Node.js 24 LTS or newer — Atomic requires the latest Node LTS runtime. Check with
node --version. - A package manager — use npm (included with Node), pnpm, Yarn, or Bun. Use Bun 1.3.14+ for Bun installs or workflow-authoring examples.
- Model-provider access — Use
/loginafter startup. Supports provider subscriptions and APIs.
Install
With npm:
npm install -g @bastani/atomicWith pnpm:
pnpm add -g @bastani/atomicWith bun:
bun add -g @bastani/atomicAtomic does not require package install scripts. If you want to disable dependency lifecycle scripts during the Atomic install, you can add --ignore-scripts to the install command.
1. Workflows
Workflows define the outer loop: inputs, steps, branches, parallelism, retries, checks, artifacts, and human review gates. Each stage runs an Atomic coding-agent session with the model provider you configured.
| Workflow | What it does | Example input |
|---|---|---|
goal | Focused workflow for small-to-medium changes when you can name the scope, exact desired outcome, and validation in the objective. It runs bounded worker turns, stores receipts in a goal ledger, requires reviewer quorum before completion, and stops as complete, blocked, or needs_human. | /workflow goal objective="Update the CLI docs for --json, include one example, run the docs build, and finish when it passes" |
ralph | Heavier plan → orchestrate → simplify → discover → review loop for larger migrations, broad refactors, multi-package changes, and spec-to-PR work. It writes RFC-style specs, delegates implementation through sub-agents, iterates on reviewer feedback, and prepares a PR report. | /workflow ralph prompt="Plan, implement, review, and prepare a PR for specs/2026-03-rate-limit.md" |
deep-research-codebase | Repo-wide research for broad, cross-cutting questions. It scouts the codebase, runs parallel specialist waves, aggregates findings, and writes durable research artifacts under research/. Prefer /skill:research-codebase for a focused subsystem or question. | /workflow deep-research-codebase prompt="How do payment retries work end to end?" |
open-claude-design | End-to-end design generation: discovers your design system, generates from a prompt, refines with feedback, and exports a handoff directory. | /workflow open-claude-design prompt="Team activity feed" reference=./mocks/feed.png output_type=prototype |
| _author your own_ | Anything outside the built-ins: issue-to-PR, review-to-merge, migration, triage, release, compliance, or team-specific review pipelines. Describe the process in natural language and Atomic can scaffold a defineWorkflow() file with typed CLI flags. | _"Create a reusable workflow that takes an issue, writes a plan, creates a branch, runs implementation and review stages, runs tests and lint, then stops for approval before final output."_ |
Run /workflow list to see installed workflows and /workflow inputs <name> for input schemas. /workflow status <id>, /workflow connect <id>, and /workflow resume <id> manage running or paused runs. Runnable references live in packages/coding-agent/examples/.
Built on Pi, extended for workflows
Atomic is the Atomic-branded fork of Pi's coding-agent CLI. The published @bastani/atomic package bundles first-party workflow, sub-agent, MCP, web-access, and intercom extensions.
That means Atomic is itself the coding-agent runtime: the selected model gets file editing, shell, write/edit tools, MCP, skills, workflows, and sub-agent capabilities inside Atomic. Atomic connects to model providers directly through API keys or supported subscription login.
Pi gives Atomic a mature, extensible harness. Atomic adds process-as-code for coding-agent work: workflow files, review gates, artifacts, resumable runs, and multi-stage execution.
---
How is Atomic different from Claude Code Dynamic Workflows?
Claude Code Dynamic Workflows and Atomic are trying to solve a similar class of problem: important software engineering work is too large for one agent pass, so the system should split the job into stages, run agents in parallel, verify the result, and keep enough state to finish long-running work.
The difference is where control lives.
| Dimension | Atomic | Claude Code Dynamic Workflows |
|---|---|---|
| Core idea | Open-source, repo-native workflow automation for coding agents. You can run built-ins, tell the coding agent to use a workflow for a task, describe new workflows in natural language for Atomic to scaffold dynamically, or version them as explicit TypeScript files. | Claude dynamically creates orchestration scripts for a task and fans work out to many parallel Claude subagents. |
| Best fit | Teams that want repeatable software engineering workflows they can inspect, version, extend, and run across providers. | Claude Code users who want Claude to decide when a task needs a larger dynamic workflow and orchestrate it automatically. |
| Workflow control | The process is explicit: stages, inputs, handoffs, retries, artifacts, model choices, and human gates are part of the workflow definition. | The process is generated dynamically by Claude for the current task, with confirmation before the first workflow run. |
| Models | Model-agnostic. Atomic connects directly to supported API-key and subscription providers, and workflows can use model fallback chains. | Claude-first. Availability is tied to Claude Code, Claude plans, and Anthropic-supported API/cloud channels. |
| Extensibility | Built on Pi extensions: add tools, TUI, MCP, web access, intercom, skills, prompt templates, themes, custom providers, and packaged workflows. | Optimized for Claude Code's built-in dynamic orchestration experience rather than an open extension SDK you own in-repo. |
| Artifacts and auditability | Research docs, specs, logs, transcripts, reviewer notes, check output, and final summaries can live in the repo or workflow run directory. | Progress is saved and resumable, but the orchestration is primarily a Claude Code runtime behavior. |
| Cost/scale posture | You choose the graph and concurrency. Atomic can be small and deterministic, or broad when you intentionally design a larger workflow. | Designed for large fan-outs, including tens to hundreds of subagents; Anthropic notes it can consume substantially more tokens than a typical Claude Code session. |
FAQ
高质量的开源MCP工具，支持动态工作流
- 需要让 Claude / Cursor 操作本地工具的 AI 工程师
- 构建多智能体协作系统的 Agent 开发者
- 配置 MCP 服务器时建议使用 stdio 传输 + JSON-RPC，避免暴露公网
- Agent 任务先做 dry-run 验证工具调用链，再开启自主执行
- API key 直接提交到 git 仓库（请用 .env 并加入 .gitignore）
- MCP 配置路径拼错或权限不足，重启 Claude Desktop 才生效
- CLI：直接 npm install -g / pip install，命令行调用
- 云端托管：可放在 Vercel / Railway / Fly.io 等 PaaS 平台
⚡ 核心功能
- 通过标准 MCP 协议与 Claude、Cursor 等主流 AI 客户端深度集成
- 提供结构化工具调用接口，显著降低 AI 集成复杂度
- 支持 Claude Desktop 和 Claude Code 无缝接入，开箱即用
- 可与其他 MCP 工具组合叠加，构建完整 AI 工作站
- 轻量无侵入设计，不影响现有系统架构
- 需要让 Claude / Cursor 操作本地工具的 AI 工程师
- 构建多智能体协作系统的 Agent 开发者
- 配置 MCP 服务器时建议使用 stdio 传输 + JSON-RPC，避免暴露公网
- Agent 任务先做 dry-run 验证工具调用链，再开启自主执行
- API key 直接提交到 git 仓库（请用 .env 并加入 .gitignore）
- MCP 配置路径拼错或权限不足，重启 Claude Desktop 才生效
👥 适合人群
🎯 使用场景
- 在 Claude Desktop 对话中直接调用本地工具，实现 AI 与系统的深度联动
- 通过自然语言驱动复杂的多步骤自动化任务，代替繁琐手动操作
- 将多个 MCP 工具组合使用，构建个人专属 AI 工作站
⚖️ 优点与不足
- +标准化 MCP 协议，生态互联性强
- +与 Claude 官方生态无缝对接
- +即插即用，配置简单快捷
- −依赖 Claude 客户端，非 Claude 用户无法使用
- −MCP 协议仍在持续演进，接口可能变更
- −需要一定的配置步骤
该工具使用 NOASSERTION 协议，商用场景请仔细阅读协议条款，必要时咨询法律意见。
AI Skill Hub 为第三方内容聚合平台，本页面信息基于公开数据整理，不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后，再部署至生产环境，并做好必要的安全评估。
📄 NOASSERTION — 请查阅原始协议条款了解具体使用限制。
🔗 相关工具推荐
❓ 常见问题 FAQ
AI Skill Hub 点评：原子工作流 的核心功能完整，质量良好。对于Claude Desktop / Claude Code 用户来说，这是一个值得纳入个人工具库的选择。建议先在非生产环境试用，再逐步推广。
| 原始名称 | atomic |
| 原始描述 | 开源MCP工具：Dynamic workflows for software engineering with Pi extensions, custom models, MC。⭐239 · TypeScript |
| Topics | mcptypescript动态工作流 |
| GitHub | https://github.com/bastani-inc/atomic |
| License | NOASSERTION |
| 语言 | TypeScript |
收录时间：2026-06-05 · 更新时间：2026-06-05 · License：NOASSERTION · AI Skill Hub 不对第三方内容的准确性作法律背书。