兰托 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。
兰托 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。
兰托 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:cargo install(推荐) cargo install lantor # 方式二:从源码编译 git clone https://github.com/chenzl25/lantor cd lantor cargo build --release # 二进制在 ./target/release/lantor
# 查看帮助 lantor --help # 基本运行 lantor [options] <input> # 详细使用说明请查阅文档 # https://github.com/chenzl25/lantor
# lantor 配置说明 # 查看配置选项 lantor --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 LANTOR_CONFIG="/path/to/config.yml"
<p align="center"> <img src="docs/assets/lantor-banner.png" alt="Lantor - AI Agent Workspace" width="820" /> </p>
- Workspace primitives — channels, DMs, threads, mentions, search, tasks, reminders, artifacts, and attachments. - Local supervisor — durable inbox dispatch, queued runs, stop/retry, process lifecycle, run logs, and structured event ingestion. - Agent collaboration — task claiming, task/thread handoff, progress activity, generated artifacts, and per-agent local memory. - Desktop + mobile access — native macOS app plus a trusted-network web UI served by the same local process and SQLite database.
Lantor is a native macOS desktop app. Install Node 20+ and Rust first:
brew install node
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y
source "$HOME/.cargo/env"
If Rust or Tauri reports missing Apple compiler or linker tools, run xcode-select --install and launch again.
Clone and launch the app:
git clone https://github.com/chenzl25/lantor.git
cd lantor
npm install
npm run tauri:dev
When the desktop app opens, add your first agent:
1. Install and sign in to the CLI runtime you want to use. You only need the runtime for the agents you plan to run:
# Codex
npm install -g @openai/codex
codex
# Claude Code
npm install -g @anthropic-ai/claude-code
claude
2. In Lantor, create an agent, choose Codex or Claude, and point it at a workspace directory. 3. Mention the agent in a channel, DM it directly, or create a task. Lantor records the work item, wakes the local CLI runtime, and routes the response back into the right thread.
SQLite state lives at ~/Library/Application Support/Lantor/lantor.sqlite, attachments live under ~/Library/Application Support/Lantor/attachments/, and migrations run automatically on every start.
Defaults work out of the box. The two settings most users care about:
| Variable | Default | Purpose |
|---|---|---|
LANTOR_DATABASE_URL | sqlite://~/Library/Application Support/Lantor/lantor.sqlite | SQLite database URL. |
LANTOR_WEB_BIND | 0.0.0.0:8787 | Web UI bind. Use 127.0.0.1:8787 for loopback only, or off to disable. |
Advanced options — attachment paths, web public URL, web bundle override, warm Codex rotation — are in docs/configuration.md and .env.example.
高质量的AI工作流项目
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。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
经综合评估,兰托 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 | lantor |
| 原始描述 | 开源AI工作流:Local-first, private AI agent workspace for Codex and Claude. Own your context a。⭐22 · Rust |
| Topics | AIRust工作流 |
| GitHub | https://github.com/chenzl25/lantor |
| License | Apache-2.0 |
| 语言 | Rust |
收录时间:2026-05-26 · 更新时间:2026-05-30 · License:Apache-2.0 · 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