easylink-ai-open/tiri-agent — explained in plain English
Analysis updated 2026-05-18
Run a personal AI assistant locally that keeps all chat history and data on your own machine.
Give an AI assistant the ability to read, write, and run commands in a sandboxed local or Docker workspace.
Connect an AI assistant to Telegram, WeChat, or Feishu so it can chat with you through those apps.
| easylink-ai-open/tiri-agent | razr001/align-dev | xiaolongbao0709/ai-virtual-phone | |
|---|---|---|---|
| Stars | 257 | 257 | 258 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | easy | moderate |
| Complexity | 4/5 | 2/5 | 3/5 |
| Audience | developer | developer | general |
Figures from each repo's GitHub metadata at analysis time.
Requires a Python virtual environment, the separate agent-runtime package, and an AI provider API key.
Tiri Agent is a personal AI assistant that runs on your own machine rather than in someone else's cloud. It is built on top of a separate core project called agent-runtime, and this repository turns that core into something you can actually use day to day: a command line chat, a set of backend APIs, and a web interface built with Next.js and React. The project keeps its data local. It stores chat sessions, message history, settings, and other records in a SQLite database on your own computer, defaulting to a folder called .tiri in your home directory. You configure which AI model it talks to by setting a handful of environment variables, such as which provider API to use, the model name, and an API key, then start it either as a command line chat tool or as a running server with both a backend and a web page. Beyond chat, Tiri can act on your behalf using sandbox tools that read files, write files, list files, and run shell commands. These can run directly on your machine or inside a Docker container for extra isolation, with one container reused per chat session so state persists across a conversation. It also supports connecting to messaging platforms. The project includes adapters for Telegram, Weixin or WeChat, and Feishu or Lark, so the assistant can send and receive messages through those channels, not just through the local CLI or web shell. Setting it up involves creating a Python virtual environment, installing both this repository and its companion agent-runtime package, and configuring your chosen AI provider's credentials before running the CLI or server commands. The project is aimed at developers comfortable working with Python and command-line tools rather than a plug-and-play consumer app. Tiri Agent is released under the GNU Affero General Public License version 3.
A local-first personal AI assistant with CLI chat, a web interface, file and shell tools, and messaging platform integrations.
Mainly TypeScript. The stack also includes Python, TypeScript, Next.js.
AGPL-3.0 requires that if you modify this and let others use it over a network, you must share your source code too.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.