开源AI工作流 是 AI Skill Hub 本期精选Agent工作流之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
开源AI工作流 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。
开源AI工作流 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。
# 方式一:npm 全局安装 npm install -g mergewatch.ai # 方式二:npx 直接运行(无需安装) npx mergewatch.ai --help # 方式三:项目依赖安装 npm install mergewatch.ai # 方式四:从源码运行 git clone https://github.com/santthosh/mergewatch.ai cd mergewatch.ai npm install npm start
# 命令行使用
mergewatch.ai --help
# 基本用法
mergewatch.ai [options] <input>
# Node.js 代码中使用
const mergewatch.ai = require('mergewatch.ai');
const result = await mergewatch.ai.run(options);
console.log(result);
# mergewatch.ai 配置说明 # 查看配置选项 mergewatch.ai --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 MERGEWATCH.AI_CONFIG="/path/to/config.yml"
<p align="center"> <img src="assets/mergewatch-wordmark.svg" alt="mergewatch" height="48" /> </p>
<p align="center"> <strong>Open-source AI code reviews for every pull request.</strong> </p>
<p align="center"> <a href="https://github.com/santthosh/mergewatch.ai/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/santthosh/mergewatch.ai/ci.yml?style=flat-square&label=tests" alt="Tests"></a> <a href="https://github.com/santthosh/mergewatch.ai/actions/workflows/docker-publish.yml"><img src="https://img.shields.io/github/actions/workflow/status/santthosh/mergewatch.ai/docker-publish.yml?style=flat-square&label=docker" alt="Docker"></a> <img src="https://img.shields.io/badge/coverage-85%25-brightgreen?style=flat-square" alt="Coverage"> <a href="https://github.com/santthosh/mergewatch.ai"><img src="https://img.shields.io/github/stars/santthosh/mergewatch.ai?style=flat-square" alt="Stars"></a> <a href="https://github.com/santthosh/mergewatch.ai/issues"><img src="https://img.shields.io/github/issues/santthosh/mergewatch.ai?style=flat-square" alt="Issues"></a> <a href="LICENSE"><img src="https://img.shields.io/badge/license-AGPL--3.0-blue?style=flat-square" alt="License"></a> <img src="https://img.shields.io/badge/PRs-welcome-brightgreen?style=flat-square" alt="PRs Welcome"> </p>
<p align="center"> <strong>We’re onboarding our first design partners</strong> — free through GA, direct line to the founder. <a href="https://tally.so/r/GxolrQ"><strong>Join →</strong></a> </p>
---
MergeWatch is a GitHub App that reviews every pull request with a team of specialized AI agents running in parallel. Security, bugs, style, error handling, test coverage, and comment accuracy are each reviewed independently, then an orchestrator deduplicates findings and scores merge readiness from 1 to 5.
Install it on your repos and it just works. Every PR gets a structured review comment, a Check Run in the merge box, and optionally a Mermaid architecture diagram of the changes.
Three services. One command. No cloud account required.
```bash git clone https://github.com/santthosh/mergewatch.ai.git && cd mergewatch.ai cp .env.example .env
gh release create v0.2.0 --generate-notes ```
Images are published to ghcr.io/santthosh/mergewatch and ghcr.io/santthosh/mergewatch-dashboard:
| Tag | When |
|---|---|
0.2.0 | On GitHub Release for v0.2.0 |
0.2 | On GitHub Release for v0.2.x (tracks latest patch) |
latest | On every push to main |
abc1234 | On every push to main (commit SHA) |
docker compose pull && docker compose up -d ```
Create a GitHub App first (permissions: pull_requests rw, contents r, checks rw, issues rw; events: pull_request, issue_comment, installation).
| Variable | Required | Notes |
|---|---|---|
GITHUB_APP_ID | Yes | From GitHub App settings |
GITHUB_WEBHOOK_SECRET | Yes | Set when creating the app |
GITHUB_PRIVATE_KEY | Yes* | Inline PEM with \n escaping |
GITHUB_PRIVATE_KEY_FILE | Yes* | Path to .pem file (alternative) |
GITHUB_CLIENT_ID | Yes | GitHub App OAuth credentials |
GITHUB_CLIENT_SECRET | Yes | GitHub App OAuth credentials |
NEXTAUTH_SECRET | Yes | openssl rand -base64 32 |
LLM_PROVIDER | Yes | anthropic / litellm / ollama / bedrock |
ANTHROPIC_API_KEY | If anthropic |
\*Exactly one of GITHUB_PRIVATE_KEY or GITHUB_PRIVATE_KEY_FILE is required.
See .env.example for the full list including LiteLLM, Ollama, and Bedrock options.
Drop a .mergewatch.yml in your repo root:
```yaml model: anthropic.claude-sonnet-4-20250514
该项目提供了一个开源的AI工作流,用于自动化代码审查和PR reviews,支持多种云平台和模型。虽然项目质量较高,但仍需要进一步优化和完善。
该工具使用 AGPL-3.0 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。
⚠️ AGPL 3.0 — 最严格的 Copyleft,网络服务端使用也需开源,SaaS 使用受限。
经综合评估,开源AI工作流 在Agent工作流赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | mergewatch-ai |
| 原始描述 | 开源AI工作流:AI-powered PR reviews running in your Cloud. Your models, your cloud, your rules。⭐10 · TypeScript |
| Topics | workflowai-code-reviewautomated-code-reviewbedrockclaudecode-reviewtypescript |
| GitHub | https://github.com/santthosh/mergewatch.ai |
| License | AGPL-3.0 |
| 语言 | TypeScript |
收录时间:2026-05-22 · 更新时间:2026-05-22 · License:AGPL-3.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端