shiwenwen/hope-agent

🦭 会记忆、能持续推进目标、会动态编排多 Agent 的跨端桌面 AI 助手，也可服务化常驻 NAS / 云端 | A cross-device desktop AI agent with memory, autonomous goals, dynamic workflows, and headless deployment

What it solves

Hope Agent is a local-first, desktop-oriented personal AI agent designed to move beyond simple chat interfaces. It solves the problem of AI assistants being confined to chat boxes by providing a tool that can manage long-term goals, execute complex workflows in the background, and maintain a persistent memory of the user's preferences and knowledge across different devices and platforms.

How it works

Built with Rust (ha-core) and Tauri, the project operates in three modes: a full Desktop GUI, a headless Server with a Web GUI, and an ACP (stdio) mode for IDE integration. It uses a "Goal-Workflow-Loop" mental model where goals define the desired outcome, workflows organize the execution, and loops determine when to trigger subsequent actions. It integrates with over 40 model providers and supports local model installation via Ollama.

Who it’s for

It is for individuals who want a powerful, private AI assistant that can operate their computer, manage a "second brain" of Markdown notes, and handle long-term tasks autonomously without requiring constant manual prompting.

Highlights

Autonomous Execution: Capable of controlling the desktop, browser, and running commands in a Docker sandbox with user approval.
Long-term Memory: Implements layered memory (global, project, agent) with vector retrieval and a "Dream Diary" for reflecting on user habits.
Design Space: Generates deliverable designs (web, mobile prototypes, presentations) from text or images with real-time previews.
Knowledge Space: Integrates with Obsidian and uses Markdown for a shared AI-human knowledge base.
Cross-platform Sync: Sessions and memory are shared across Desktop, Web, and various IM channels like Telegram, Discord, and Slack.