Radar check
What we found when we opened it. Every bar is one fact, with its date.
Last code pushed 27 Oct 2026, one day before this check.
Latest release v0.4.3 on 23 Aug 2026.
Public since March 2025, a little over 1 year.
2,760 stars and 296 forks.
30 commits and 5 new issues in the last 90 days.
10 contributors. 6 pull requests merged in 90 days.
No security policy file says how to report a problem. No published security advisories.
How the bars are scored
- Alive. Days since the last code push: a week or less 100, a month 90, three months 70, six months 50, a year 30, eighteen months 15, older 0.
- Releases. Age of the latest GitHub release, on a similar scale; no bar when the project publishes no releases.
- Maturity. Age of the project: eight years 100, five 90, three 70, two 55, one 40, six months 25, younger 10.
- Adoption. Stars on a log scale: 100,000 stars is 100, 10,000 is 80, 1,000 is 60, 100 is 40.
- Activity. Commits in the last 90 days: 300 or more 100, 100 85, 30 65, 10 45, any 25.
- Community. People who have contributed: 100 or more 100, 30 80, 10 60, 5 40, 2 25, one 10.
- Security. A published security policy 80, none 40; published advisories are listed, not scored.
Checked by Repo Radar on 28 Oct 2026 from the repository on GitHub.
In simple English
This tool creates a desktop companion that talks to you, controls your computer, and connects to chat apps. It features animated avatars that react to your voice and can browse the web for you. You can run it locally on your computer or host it on a server. It works with various artificial intelligence models and supports custom extensions. The software is free to download, but you need your own AI model keys.
What it is
Super Agent Party is a self-hosted artificial intelligence companion platform that combines animated desktop avatars with task automation and chat integrations. It provides a web interface, desktop app wrappers, and messaging bots for platforms like Telegram and Discord.
What it is good for
Solo builders who want an interactive artificial intelligence assistant on their desktop that can perform background tasks, manage chat channels, and run custom skills without relying on proprietary third-party desktop apps.
Summary (Key Points)from the README
Key features
- Desktop companion: Supports custom VRM models, actions and 3D scenes, and Live2D avatars can be driven through the Vtube Studio (VTS) integration.
- Task Center: Lets AI agents run advanced tasks in the background, with support for MCP and Agent Skills.
- Computer control: Combines desktop vision with a toolchain that controls the mouse, keyboard and terminal.
- Window modes: The chat interface docks to the desktop in phone-sized or capsule-sized windows.
- Multi-role group chat: Chat with several characters at once, using tavern character cards and long-term memory.
- Messaging and live-streaming bots: One-click deployment to QQ, WeChat, Feishu, DingTalk, Telegram, Discord, Slack, Bilibili, YouTube and Twitch.
- AI browser: Gives agents their own browser with automatic control.
- Extension system: Install existing extensions or build new ones, which open in a separate window or a sidebar. The README lists 17 extensions in its table.
- Developer APIs: Offers OpenAI-compatible and MCP interfaces, so agents can be connected from outside. The README says existing OpenAI code needs no changes.
- Light hardware needs: The README lists 2 CPU cores and 2GB of memory, and says the Docker version runs on a 2-core, 2G cloud server when cloud model providers are used.
Ways to install
| Platform or method | What it needs | Effort |
|---|---|---|
| Windows portable zip | Windows 10/11 or Windows Server 2025; extract and run the start script | Low |
| Windows desktop installer | Windows 10/11 or Windows Server 2025; install for the current user only | Low |
| macOS integrated package | Apple M-series chip; remove quarantine, set script permissions, then run update and start scripts | Medium |
| macOS desktop app (.dmg) | Apple M-series chip; move the app to Applications and remove quarantine | Low to medium |
| Linux AppImage | x86_64 or arm64; no installation needed | Low |
| Docker | Docker; pull and run the image, then open port 3456 in a browser | Low to medium |
| Docker Compose | Git and Docker Compose; clone the repo and start the containers | Medium |
| Lite client | Windows or macOS; a desktop wrapper for the Docker version | Low |
| Source code | uv and npm; sync dependencies, then run the dev server | High |
Important
- Browser-only Docker: The README says the Docker versions show the desktop pet only through a browser.
- Default login: The Docker Compose setup uses the username `root` and password `pass`, and the README asks users to change the password after the first login.
- Edge TTS is optional: It is off by default. It uses an undocumented Microsoft endpoint that may stop working at any time, and commercial use may breach Microsoft's terms.
- Browser control limits: The README says it is meant for accessibility and must not be used on banking, payment or highly confidential pages.
- Licensing: The default license is AGPLv3. Closed-source commercial use requires a separate commercial license from the project administrator.
- Third-party model output: The README places responsibility for generated content on model providers and users, and disclaims warranties.
Written by an AI model from the project’s README; not part of the review, and it can be wrong.
- Supports custom animated avatars including VRM models and Vtube Studio integration.
- Deploys easily as a desktop application, Docker container, or source installation.
- Connects directly to messaging platforms like Telegram, Discord, and Slack.
- Requires macOS users to run terminal commands to bypass quarantine restrictions.
- Memory usage depends heavily on the chosen artificial intelligence models and background tasks.
- AGPL-3.0 licence requires releasing source code if you modify and host it publicly as a service.
Proprietary desktop assistant applications and manual chat bot integrations
Ollama
- Significant-Gravitas/AutoGPT188k stars
An autonomous agent framework that executes complex multi-step tasks using large language models.
- lobehub/lobehub84k stars
A chat interface for language models featuring custom plugins and character personas.
- vllm-project/vllm96k stars
A serving engine that makes one graphics card answer many requests at once.
- infiniflow/ragflow93k stars
A question-answering system built on a document collection, with the sources shown beside every answer.
- basicmachines-co/basic-memory4.3k stars
Gives an AI assistant a memory that is stored as plain text files you can read.
git clone https://github.com/heshengtao/super-agent-party.git cd super-agent-party uv sync npm install npm run dev