5ire

5ire is an MCP client/server that exposes an AI-assisted MCP server toolset. It enables a sleek AI assistant that can access tools, manage prompts, knowledge bases, and retrieval-augmented generation capabilities through MCP-compliant endpoints. Once running, you can interact with the 5ire MCP server to issue tool-enabled queries, fetch system information, read documents, perform searches across conversations, and leverage the local knowledge base integration for multilingual vectorization and document parsing. The server is designed to plug into a broader MCP ecosystem, making it easy to discover and connect additional MCP servers via the marketplace, and to share your own tool integrations with others. Whether you want to query your local files, run database lookups, or perform RAG-style retrieval over documents, 5ire provides a unified interface to access these capabilities through the MCP protocol.

Prerequisites:

• Node.js installed on your system
• Python installed on your system
• uv (Python package manager) available if you plan to use Python-based tools

Option A: Quick start with npx (no installation required beyond npm/yarn)

1. Ensure Node.js and npm are installed.
2. Run the MCP server via npx:

npx -y 5ire

Option B: Install via npm package (permanent installation)

1. Install the package globally (or view local installation options):

npm install -g 5ire

2. Start the server using the package command (adjust path if needed):

node path/to/server.js

Option C: If you prefer a Python/uv-based setup for a Python runtime environment (if provided as a Python package in your workflow)

1. Ensure uv is installed and Python is available
2. Install and run according to the Python-based instructions in the repository (example placeholder):

python -m uvx install 5ire
uvx run 5ire

Note: The exact install command may vary by release; refer to the repository’s installation guide for the precise steps and package names.

Tips and common considerations:

• The 5ire server relies on MCP protocols to expose tools and prompts to LLMs; ensure your environment has network access to any external data sources the tools may query.
• If you encounter notarization or security prompts on macOS, you may need to configure Apple-specific notarization environment variables (APPLE_TEAM_ID, APPLE_ID, APPLE_ID_PASS).
• The MCP marketplace allows discovering and sharing MCP servers; consider exploring the marketplace to find compatible tools that suit your needs.
• If you integrate 5ire into a web app, you can use the one-click server installation guide to embed the server setup into your site.
• For multilingual parsing and local vectorization, the server’s local knowledge base features rely on an embedded model (e.g., bge-m3); ensure the model is installed and accessible in your environment.
• If you run into tool invocation errors, verify that Python and Node runtimes are correctly configured and that any required environment variables (such as paths to tool binaries or API keys) are set appropriately.
• When using the MCP tools, you can access file systems, databases, remote data sources, and document formats (docx, xlsx, pptx, pdf, txt, csv) through the local knowledge base and tool integrations.