Happy Friday Lite

August 24, 2026 · View on GitHub

中文

An Electron + Vue 3 desktop personal knowledge assistant that combines an AI agent, RAG knowledge retrieval, rich-text notes, and a calendar. All data is stored locally with privacy as a priority.

✨ Key Features

🤖 Friday AI Assistant

  • Built with DeepAgent SDK and LangChain
  • 12+ built-in tools: knowledge-base search, note and schedule management, file operations, shell execution, Python REPL, HTTP requests, web scraping, JSON processing, calculator, and more
  • Human-in-the-Loop (HITL) support for approval before sensitive operations
  • SKILL system that loads Markdown skill descriptions from the SKILL/ directory
  • Subagents and cross-session memory powered by SQLite and InMemoryStore

📚 Knowledge Base

  • Personal, local, and agent knowledge bases
  • Full RAG pipeline: document loading -> parent-child chunking -> embedding -> vector search
  • Unified Zvec vector storage, separated by kb_type
  • File watching, scheduled incremental indexing, and manual updates
  • PDF, Word, Excel, Markdown, HTML, EPUB, and plain-text support

📝 Notes

  • TipTap rich-text editor
  • Syntax-highlighted code blocks, tables, task lists, images, links, and more
  • AI Fill-in-the-Middle (FIM) completion
  • Version history and diff comparison

📅 Schedule

  • Schedule-X calendar view with lunar calendar support
  • Create and update schedules with natural language through Friday

🎨 More

  • Light, dark, and system themes
  • Simplified Chinese and English localization
  • Automatic data backups and a multi-tab interface

🖼️ Screenshots

Friday AI Assistant

Friday assistant home

Friday chat and code execution

Knowledge Base

Knowledge base file cards

Notes

Rich-text note editor

Schedule

Schedule calendar

Automated Tasks

Automated tasks

DeepSeek Harness

DeepSeek Harness

Settings

Settings and feature overview

🚀 Quick Start

Requirements

  • Node.js >= 20 (22 recommended)
  • npm >= 10
  • macOS, Windows, or Linux

Install Dependencies

npm install --legacy-peer-deps

Run in Development

# Frontend only
npm run dev

# Electron + frontend (starts Electron automatically)
npm run electron:dev

Build a Release Package

# Build an installer for the current platform
npm run electron:build

# Build for arm64
npm run electron:build:arm64

Build artifacts are written to release/.

Configure Python under Settings -> General -> Python Environment. Use Auto Detect to find an installed interpreter, or select its executable manually. The AI assistant can install missing dependencies from python/requirements.txt using pip/pip3.

📦 Download

Download the installer for your platform from Releases:

PlatformArtifact
macOS (Apple Silicon)*-mac-arm64.dmg
Windows*-win-x64-setup.exe
Linux*-linux-*.AppImage

🛠️ Tech Stack

LayerTechnology
Desktop frameworkElectron 42
Frontend frameworkVue 3 + Vite 6
State managementPinia
RoutingVue Router
InternationalizationVue I18n
Rich-text editorTipTap
CalendarSchedule-X
AI agentDeepAgents + LangChain
Vector databaseZvec
Local databaseSQLite (sql.js)
Python runtimeBundled portable Python for the Agent python_repl tool

📂 Project Structure

happy-friday-lite/
├── main.js                  # Electron main-process entry
├── preload.cjs              # Preload script
├── src/                     # Vue frontend source
│   ├── components/          # Shared components (chat / layout)
│   ├── views/               # Page views
│   │   ├── friday/          # Friday assistant
│   │   ├── knowledge/       # Knowledge base
│   │   ├── note/            # Notes
│   │   ├── schedule/        # Schedule
│   │   ├── history/         # Version history
│   │   └── settings/        # Settings
│   ├── store/               # Pinia state modules
│   ├── i18n/                # Localization resources
│   └── router/              # Router configuration
├── src-electron/            # Electron backend
│   ├── agent/               # Agent core (tools / skills / subagents / permissions / memory)
│   ├── rag/                 # RAG pipeline (loading / chunking / embedding / retrieval)
│   ├── db.js                # SQLite database
│   ├── llm.js               # LLM adapter
│   └── python-env.js        # Python runtime environment
├── python/                  # Python runtime download scripts
├── scripts/                 # Build helper scripts
└── .github/workflows/       # CI build workflows

🔧 Configure an AI Model

On first use, go to Settings -> Model Configuration and enter an API Base URL (for example, https://api.openai.com/v1), API key, and model name.

Any OpenAI-compatible provider is supported, including OpenAI, DeepSeek, Qwen, Zhipu AI, Kimi, and Doubao.

🤝 Contributing

Issues and pull requests are welcome. Please ensure the project builds successfully, changes follow the existing code style, and documentation is updated with new features.

📄 License

This project is released under the PolyForm Noncommercial License 1.0.0.

  • ✅ Allowed: personal use, learning and research, modifying the code, and distributing the source code
  • ❌ Prohibited: commercial use in any form
  • ⚠️ Required: retain the original author's attribution and copyright notice when using or distributing the project

Read LICENSE for details.

🙏 Acknowledgments

📧 Contact

For questions or suggestions, open an Issue or email chenjie.plus@qq.com.