Playa-Cyrene/Cyrene-Agent

一款开源、基于昔涟开发的桌面 AI 智能体/An open-source desktop AI agent based on Cyrene.

What it solves

Cyrene-Agent provides a personalized, interactive AI desktop companion based on the character "Cyrene" from Honkai: Star Rail. It integrates LLM-driven conversation with practical desktop automation, allowing the character to assist with work, coding, learning, and daily tasks while maintaining a consistent persona.

How it works

Built on Electron and TypeScript, the project uses a custom "CyreneHarness" engine to manage AI agent loops, tool execution, and memory. It features a Live2D character that resides on your desktop, responding to inputs with animations and expressions. The agent supports four modes: Chat (pure conversation), Work (web/file/document tools), Code (LSP-based coding assistance), and Learn (Obsidian-linked study support). It utilizes a multi-layered memory system (DMAE) to track user interactions and preferences over time.

Who it’s for

It is designed for Windows users who want a highly interactive, character-based AI assistant that can perform real-world tasks like web browsing, file management, and code analysis while providing a more immersive, personality-driven desktop experience.

Highlights

  • Live2D Desktop Presence: Features a fully animated character with state-aware expressions and interactions.
  • Versatile Agent Modes: Dedicated modes for chatting, professional work, coding (with LSP support), and learning.
  • Advanced Memory System: Uses a multi-tiered memory engine to maintain long-term context and user relationships.
  • Extensible Plugin Architecture: Allows developers to create custom tools, new chat channels, and UI extensions using a provided SDK.
  • Multi-Channel Support: Can be accessed via desktop, Feishu, WeChat (iLink), and QQ.

Related

  • Project
    RachelForster/Shinsekai

    Shinsekai is a desktop assistant for visual‑novel style RPGs that uses LLMs to generate character dialogue, synchronises portrait emotions, and optionally adds TTS/ASR and image generation. All components are configured via a React settings UI, can be swapped (different LLMs, TTS engines, ComfyUI, etc.), and are extensible through a plugin SDK or the Model Context Protocol. The project provides ready‑to‑run install scripts, a plugin registry, and a Tauri‑based native desktop wrapper.

  • Project
    zhayujie/CowAgent

    CowAgent is an open‑source, self‑hosted AI assistant that plans tasks, runs tools, keeps a personal knowledge base, and can operate as a team of cooperating agents. It supports many LLM providers, multimodal input, dozens of chat channels (Telegram, Slack, WeChat, etc.), a plug‑in skill system, and a three‑tier memory with nightly “Deep Dream” distillation. Install via a one‑line script or Docker, manage everything through a web console, and extend the system through the public Skill Hub or custom tool integrations.

  • Project
    deepagent-ltd/deepagent-code

    DeepAgent Code (DAC) is a desktop‑plus‑terminal workspace that lets you use an LLM as a coding assistant for multi‑step tasks. It offers persistent sessions, inspectable project memory, sub‑agents for independent jobs, a managed browser, and optional instant‑messaging integration, all configurable with any compatible LLM provider. Licensed under AGPL‑3.0.

  • Project
    modelscope/ms-agent

    MS‑Agent is an open‑source Python framework that lets you build, run, and manage autonomous AI agents. It provides a customizable harness for planning, tool execution, permission checks, and feedback, with built‑in context‑compression, long‑term memory, and DAG‑style multi‑agent workflows. Agents can be used via a Web UI, terminal UI, CLI, or embedded SDK, and the ecosystem includes reusable “skills”, multimodal input, and ready‑made applications for research, code generation, video creation, and finance.

  • Project
    microsoft/RD-Agent

    RD‑Agent is an open‑source Microsoft framework that lets LLM‑driven agents automatically propose ideas (R‑agent) and implement them (D‑agent) across data‑centric R&D tasks such as quantitative‑finance factor modelling, Kaggle‑style data‑science pipelines, and research‑paper code generation. It tops the MLE‑bench benchmark, supports any LiteLLM‑compatible model, and provides Docker‑based demos and a web UI for interactive tracing.