经 AI Skill Hub 精选评估,AI工作流可视化 获评「强烈推荐」。这款Agent工作流在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 8.0 分,适合有一定技术背景的用户使用。
AI工作流可视化 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。
AI工作流可视化 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 argus # 方式二:npx 直接运行(无需安装) npx argus --help # 方式三:项目依赖安装 npm install argus # 方式四:从源码运行 git clone https://github.com/duytrandt04-afk/argus cd argus npm install npm start
# 命令行使用
argus --help
# 基本用法
argus [options] <input>
# Node.js 代码中使用
const argus = require('argus');
const result = await argus.run(options);
console.log(result);
# argus 配置说明 # 查看配置选项 argus --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 ARGUS_CONFIG="/path/to/config.yml"
Website: getargus.org
The hook control center for AI coding agents. Hooks are how you govern what Claude Code and Codex can do — block dangerous commands, protect secrets, enforce branch policy, get notified. But managing them means hand-editing JSON, testing them means waiting for a live agent to misbehave, and good scripts are scattered across a thousand gists. Argus fixes all three, locally:
- Hook management — a config editor with one-click presets for Claude Code and Codex. No JSON surgery; argus-managed entries are tagged and reversible. - Hook simulator — run any hook command or script against a realistic synthetic payload for any event type, and inspect stdout/stderr/exit code/duration before an agent ever fires it. The missing debugger for the hook ecosystem. - Public script collection — my-custom-hook-scripts/ ships battle-tested, zero-dependency guardrails free for everyone: dangerous-command blocker, secrets protection, branch guard, auto-format with lint feedback, prompt-injection scanner, webhook notifications, and more. Every script works with Claude Code and Codex.
Backing it up: a live observability layer — every hook payload is normalized to a canonical event model, persisted to SQLite, and streamed to a real-time dashboard (event feed, session explorer, usage and cost stats). You see your hooks — and your agents — actually working. No cloud, no telemetry, your data stays local.
curl -fsSL https://raw.githubusercontent.com/duytrandt04-afk/argus/main/uninstall.sh | bash
Stops the server, removes binaries and scripts, unwires hooks from ~/.claude/settings.json, and optionally deletes your data.
curl -fsSL https://raw.githubusercontent.com/duytrandt04-afk/argus/main/install.sh | bash
Requirements: Node.js 18+, curl, tar — no Go or pnpm needed. The installer downloads a pre-built binary for your OS and arch, wires the Claude CodeSessionStarthook, and placesargusin~/.argus/bin.
Open http://127.0.0.1:10804 after your next Claude Code or Codex session starts.
Then follow docs/quickstart.md to verify your first event.
高质量的AI工作流可视化项目,值得关注
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。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
AI Skill Hub 点评:AI工作流可视化 的核心功能完整,质量优秀。对于自动化工程师和运维人员来说,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 | argus |
| 原始描述 | 开源AI工作流:Local-first observability dashboard for AI coding agent activity. Provide space 。⭐11 · TypeScript |
| Topics | AI工作流可视化仪表盘 |
| GitHub | https://github.com/duytrandt04-afk/argus |
| License | MIT |
| 语言 | TypeScript |
收录时间:2026-06-12 · 更新时间:2026-06-12 · License:MIT · 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