Hello everyone, I'm Programmer Wanfeng,实战 various AI projects.

March 28, 2026, Feishu officially open-sourced Lark CLI - a command-line tool designed specifically for AI Agents. With it installed, your AI (like Claude Code, Cursor) can directly control Feishu's calendar, documents, bitable, messages and all core functions.

What does this mean? Your AI assistant can finally truly "enter" your office system.

1. What is Feishu CLI?

One Sentence Explanation

Feishu CLI is Feishu's officially provided command-line interface, encapsulating Feishu open platform's 2500+ APIs into 200+ commands, allowing AI Agents to directly operate Feishu through natural language instructions.

Core Features

FeatureDescription
🚀 AI Agent NativeBuilt-in 19 Agent Skills, mainstream LLMs need no extra learning
📦 200+ CommandsCover message, document, calendar, email and other 11 major business domains
🔓 Open Source FreeMIT license, can be used commercially, can be secondary developed
🔒 Safe and ReliableOfficial product, data security guaranteed

Who is it for?

  • Heavy Feishu users (enterprises, teams)
  • People who want to automate Feishu office workflows with AI
  • Developers (can develop custom tools based on CLI)

2. Install Feishu CLI

Environment Requirements

  • macOS / Linux / Windows
  • Node.js 16+ (recommended) or Go 1.20+

Installation Methods

Method 1: npm install (recommended)

1
npm install -g @larksuite/lark-cli

Method 2: Homebrew (macOS)

1
brew install larksuite/tap/lark-cli

Method 3: Direct download

1
2
3
4
5
6
7
8
9
# macOS
curl -L https://github.com/larksuite/lark-cli/releases/latest/download/lark-cli-darwin-amd64 -o lark-cli
chmod +x lark-cli
sudo mv lark-cli /usr/local/bin/

# Linux
curl -L https://github.com/larksuite/lark-cli/releases/latest/download/lark-cli-linux-amd64 -o lark-cli
chmod +x lark-cli
sudo mv lark-cli /usr/local/bin/

Verify Installation

1
2
lark-cli --version
# Output: lark-cli version 1.0.0

3. Configure Feishu CLI

Step 1: Create Feishu App

  1. Visit Feishu Open Platform
  2. After logging in, click "Create Enterprise Self-built App"
  3. Fill in app name (e.g.: AI Office Assistant)
  4. Record App ID and App Secret

Step 2: Configure Permissions

In app backend, enable the following permissions:

  • contact:contact:readonly (read contacts)
  • im:chat:readonly (read group chat info)
  • im:message:send (send messages)
  • docs:document:readonly (read documents)
  • docs:document:write (edit documents)
  • calendar:calendar:readonly (read calendar)
  • calendar:calendar:write (edit calendar)
  • bitable:bitable:app:readonly (read bitable)

🎓 AI 编程实战课程

想系统学习 AI 编程？程序员晚枫的 AI 编程实战课 帮你从零上手！