OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Lilium AI

Lilium AI is an open-source TypeScript framework for a self-hosted personal AI agent that connects language models to browser automation, shell execution, filesystem access, scheduling, memory and messaging channels.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Lilium AI — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Lilium AI when a natural-language agent needs to combine model reasoning with browser, shell or file actions on a desktop or server, particularly when self-hosting or local Ollama models are relevant.

Who it is for

  • Developers extending an agent through JavaScript plugins
  • Users who want self-hosted computer automation
  • Teams or individuals connecting an agent to supported messaging channels

Documented uses

  • Automate browser navigation, searches, clicks and data extraction
  • Run shell commands and inspect or manipulate local files
  • Schedule recurring workflows such as web monitoring or reports
  • Transcribe voice messages and save summaries to local Markdown files
  • Expose agent interactions through channels including WeChat, Telegram and Discord
Before you choose

Requirements & limits

The repository is public and identified as MIT-licensed. Setup is self-hosted and requires Node.js, pnpm, model configuration and any needed channel credentials.

  • Choose the model backend: the repository documents OpenAI and Ollama, with Ollama configured through a local base URL.
  • Account for the authority granted by shell execution, filesystem I/O and browser automation.
  • Channel-specific setup varies, and the repository does not supply full configuration details for every named integration.
  • A dedicated test account is recommended by the project for some platform automations because of account-restriction risk.
  • macOS users may need accessibility permissions for Wechaty or Puppeteer workflows.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a self-hosted Lilium AI agent

Install the framework, configure a model and channels, and start the agent.

Before you start

  • Node.js v18 or higher
  • pnpm
  • An OpenAI API key or an Ollama base URL
  • Channel tokens for any enabled messaging channels
  • Python 3 only if using the macOS permission-checking script

Steps

  1. Clone the repository and enter its directory.

  2. Run pnpm install to install dependencies.

  3. Copy config.example.json to config.json.

  4. Edit config.json with the LLM settings, channel settings and a changed administrator password.

  5. Run pnpm start, or run pnpm build followed by the command below.

    Documented command
    node dist/index.js

Expected outcome

A locally running Lilium AI agent with the configured model and enabled channels.

Guide limitations

  • The repository does not provide implementation or configuration details for every named channel.
  • Automated interactions with some platforms may conflict with their terms, and the project warns of account-restriction risk for personal accounts.
  • This recipe documents setup steps and does not independently verify execution.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
beidald/liliumai · Official homepage title · retained in its source language
How we review sourcesSuggest a correction