Sandbox
@nanbingxyz/5ire

Desktop AI assistant and MCP client

5ire is a desktop app for chatting with major model providers and connecting them to tools through MCP servers. It also adds a local knowledge base for RAG, a prompt library, bookmarks, and search across conversations.

5,345 starsβ€’430 forksβ€’TypeScriptβ€’Updated 2mo ago
Who it's for

Builders who want a desktop AI assistant that can connect to MCP servers and local knowledge bases.

What it delivers

You can keep prompts, tools, and local reference material in one assistant instead of juggling separate apps.

What it does

MCP server support

Connects to MCP servers so the assistant can use tools, system data, files, and databases.

Local knowledge base

Parses docx, xlsx, pptx, pdf, txt, and csv files and stores embeddings for local RAG.

Prompt library

Lets you create and organize reusable prompts with variables.

Usage analytics

Tracks API usage and spending so you can see what the assistant costs.

Bookmarks and search

Saves conversations and lets you search across them later.

README

Logo
(pronounced "fai-er")

A Sleek AI Assistant & MCP Client




OpenAI / Azure / Anthropic / Google / Mistral / Doubao / Grok / DeepSeek / Ollama

Twitter Β· Releases



Before to activating tools feature, ensure the following components are installed:

  • Python
  • Node.js
  • uv (Python package manager)

These components are required as they constitute the runtime environment for the MCP Server. If you don't anticipate using the tools feature immediately, you may choose to skip this installation step and complete it later when the need arises.

For detailed installation instructions, please see our Installation Guide.


πŸ’ͺ Getting involved

If you want to contribute code to 5ire or develop your own apps based on 5ire, start with the Development Setup Guide.

Visit Wiki for more details.

πŸš€ To integrate 5ire’s one-click server installation into your website, see the One-Click Server Installation Integration Guide.


πŸ’¬ Community

Join the Discord, By joining, you’ll get:

  • Faster responses to your questions and issues
  • Direct input on new features and improvements
  • A place to share ideas, tips, and feedback with other 5ire users
  • Opportunities to co-build and shape the future of 5ire together

Features

βš’οΈ Support Tools via MCP Servers

MCP is an open protocol that standardizes how applications provide context to LLMs. Think of MCP like a USB-C port for AI applications. Just as USB-C provides a standardized way to connect your devices to various peripherals and accessories, MCP provides a standardized way to connect AI models to different data sources and tools.

With tools, you can access the file system, obtain system information, interact with databases, access remote data, and more, rather than just having a simple conversation.

https://github.com/user-attachments/assets/5aa98f2b-c26d-435e-8196-73fa414066eb

We have created an open marketplace for MCP Servers. it empowers users to discover exceptional tools while offering a streamlined process for sharing their own MCP server creations.

https://github.com/user-attachments/assets/be66c30e-bb29-4dfe-9f25-8d396470ed60

πŸ’‘ Local Knowledge Base

We have integrated the bge-m3 as our local embedding model, which excels in multilingual vectorization. 5ire now supports parsing and vectorization of docx, xlsx, pptx, pdf, txt, and csv documents, enabling storage of these vectors to power robust Retrieval-Augmented Generation (RAG) capabilities locally.

Local Knowledge Base Screenshot

πŸ“ˆ Usage Analytics

By keeping track of your API usage and spending, you can gain a better understanding of how much you're spending on the API and make informed decisions to optimize your use of the service.

Usage Analytics Screenshot

✨ Prompts Library

The prompt library provides an effective way to create and organize your own prompts. These prompts are highly versatile, thanks to their support for variables.

Prompts Library Screenshot

πŸ”– Bookmarks

You can bookmark each conversation, and even if the original messages are deleted, the saved bookmarked content remains unaffected. Bookmarks Screenshot

πŸ” Quick Search

You can perform keyword searches across all conversations, quickly pinpointing the information you need. Search Screenshot

[!TIP] Since 5ire uses native dependencies, it needs to be packaged on the corresponding platform. If it is on Mac OS, you may also need to configure APPLE_TEAM_ID, APPLE_ID, and APPLE_ID_PASS for notarization to avoid security alerts.


Discover Exceptional MCP Servers

MCPSvr, a community-driven directory of MCP servers, empowers developers to discover exceptional tools while offering a streamlined process for sharing their own MCP server creations.

Files in the repo

Repository payloadβ€’28 top-level entries
  • .erb
  • .github
  • .husky
  • build
  • docs
  • drizzle
  • patches
  • public
  • scripts
  • src
  • test
  • .better-commits.json
  • .editorconfig
  • .gitattributes
  • .gitignore
  • biome.json
  • CODE_OF_CONDUCT.md
  • DEVELOPMENT.md
  • drizzle.config.ts
  • INSTALLATION.md
  • LICENSE
  • package-lock.json
  • package.json
  • postcss.config.js
  • README.md
  • rsbuild.config.ts
  • tailwind.config.js
  • tsconfig.json

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More tools

JuliusBrussee/
caveman

πŸͺ¨ why use many token when few token do trick β€” Claude Code skill that cuts 65% of tokens by talking like caveman

105k
1 add
MemPalace/
mempalace

The best-benchmarked open-source AI memory system. And it's free.

59k
stablyai/
orca

Orca is the ADE for working with a fleet of parallel agents. Run any coding agent with your own subscription. Available on desktop, mobile and remote runtime.

66k

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io

132k

Never stop coding. Free MIT AI gateway: one endpoint, 352 providers (150+ free), 1200+ models Kimi, Claude, GPT, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 550+ contributors

64k
headroomlabs-ai/
headroom

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

71k