vixues

vixues / LeAgent

Public

Desktop AI agent with conversational chat, visual workflows, 100+ tools, skills, MCP, and multi-model support.

17
3
100% credibility
Found May 22, 2026 at 17 stars 5x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

LeAgent is an open-source desktop AI agent platform that combines a conversational assistant with visual workflow automation and over 100 built-in tools. Users can chat naturally with an AI that can read documents, run code, search the web, manage files, and execute automated workflows. The platform supports multiple AI providers (DeepSeek, OpenAI, Anthropic, etc.), runs locally on the user's machine with SQLite by default, and offers both a desktop application and web interface. It includes features like a customizable sidebar avatar, visual workflow builder with ReactFlow, a skills system for extending capabilities, and multi-channel delivery for DingTalk, Feishu, WeChat Work, and more.

How It Works

1
💬 You hear about LeAgent

A friend tells you about an AI assistant that can chat, run workflows, and handle files on your computer.

2
📥 You download and install it

You grab the installer for your computer (Windows, Mac, or Linux) and install it in a few clicks. Everything runs locally on your machine.

3
🔑 You connect your AI service

You enter your API key from DeepSeek, OpenAI, or another AI provider so LeAgent can think and respond.

4
✨ You start chatting

A friendly chat window opens. You type a question, and LeAgent thinks, uses tools, and responds with answers, files, or charts.

5
You explore different paths
📄
Process documents

Drop in a PDF or spreadsheet and ask LeAgent to summarize, extract data, or answer questions about it.

🔗
Build workflows

Drag and drop steps to create an automated pipeline that runs your tasks on a schedule.

💻
Code and create

Ask LeAgent to write code, run it in a sandbox, or help you build a small web project.

6
🤖 Your AI teammate works

LeAgent remembers your conversations, uses the right tools automatically, and handles complex tasks while you watch.

🎉 You get results

Documents processed, workflows running, code written — your tasks complete while you focus on what matters.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 17 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is LeAgent?

LeAgent is an open-source desktop AI agent built with Python that combines multi-turn conversational chat, a visual workflow editor, and over 100 built-in tools. It acts as a local AI assistant that can read documents, execute code, browse the web, manage files, and integrate with messaging platforms like DingTalk and Feishu. The project ships as a full-stack application: a FastAPI backend runs the agent runtime and tool executor, while a React frontend provides the chat interface and ReactFlow-based workflow builder.

Why is it gaining traction?

The standout hook is the "zero configuration" promise — download a desktop installer and run it without installing Python, Node, or Docker. Developers get a proper CLI (`leagent` command) for managing providers, channels, workflows, and skills from the terminal. The multi-model routing (DeepSeek, OpenAI, Anthropic, and more) means you can swap providers without rewriting prompts. The workflow engine auto-generates typed nodes from tools, so building an automation pipeline is visual rather than code-heavy.

Who should use this?

Desktop power users who want a local AI agent with file system access, document processing, and code execution should consider it. Developers evaluating open-source agent platforms for automation tasks will find the tool registry and skill system useful. Teams that need to deploy an AI assistant behind a firewall — without relying on external APIs for everything — can self-host with SQLite or PostgreSQL. The desktop installers are still in beta, so early adopters comfortable with bleeding-edge software will get the most value.

Verdict

LeAgent scores 1.0% on credibility, reflecting its infancy: 17 stars, a Chinese developer email in the package metadata, and placeholder URLs in the docs. The feature set is impressive — memory, workflows, channels, skills — but the project needs more community validation before betting production workloads on it. Worth watching, but hold off on production use until test coverage and documentation solidify.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.