IQBandit: Self-hosted AI Gateway Management Panel and Conversation Interface for OpenClaw

IQBandit is an OpenClaw gateway management panel built with Next.js and TypeScript, designed to address the challenge of balancing comprehensive functionality and user-friendly experience in the self-hosted AI infrastructure domain. It offers features such as authentication, settings management, request logging, and a conversation interface, enabling self-hosted AI gateways to have a product-grade user experience.

In the self-hosted AI infrastructure field, most local AI gateway solutions work properly but have issues like rudimentary interfaces, complex configurations, and lack of management tools, giving a "semi-finished product" impression. The IQBandit project was born to address this pain point, providing a fully functional management panel and conversation interface for OpenClaw gateways.

Architecture Design: Adopting a clear layered approach, IQBandit as the product layer handles user interface, authentication, and other experience-related functions; OpenClaw as the gateway engine layer manages request routing and model scheduling; LLM providers offer underlying AI capabilities, and each component can evolve independently. Tech Stack: Developed with Next.js App Router and TypeScript; relies on Node.js 20+ runtime environment, uses SQLite for persistent storage; implements secure authentication via JWT Cookie session mechanism; supports Docker containerized deployment.

1. Authentication and Access Control: Built-in JWT authentication system with a 7-day session validity period; management pages and APIs are protected, and there's IP-level login rate limiting (max 10 attempts in 5 minutes).
2. Dynamic Settings Management: Supports modifying configurations like chat mode, OpenClaw gateway URL, and tokens via the web interface; changes undergo strict validation.
3. Connection Testing and Diagnosis: Provides a connection test API to detect gateway health status, compatible with different deployment environments.
4. Request Logging and Observability: Records all chat requests to SQLite; logs include information like timestamp, model, status, and latency, facilitating debugging and monitoring.
5. Conversation Interface: Provides a streaming response conversation function via the /officebuilding page, with requests proxied to the OpenClaw gateway.

IQBandit maps common upstream failures to clear error messages, such as 401/403 corresponding to "Gateway token is invalid or expired", connection refused corresponding to "Gateway unreachable", etc., greatly reducing problem troubleshooting time, especially during initial deployment and configuration phases.

Deployment Options: For local development, you can install dependencies via npm, configure .env.local, generate a session key, then start; Docker deployment is recommended for production environments, supporting log directory mounting for persistent data. Security Measures: Gateway tokens are only stored on the server, with masked display on the frontend; IP-level login rate limiting; weak password warnings in production environments; upstream request timeout protection; strict settings validation mechanism.

Applicable Scenarios: Internal enterprise AI gateway management, productization of self-hosted AI infrastructure, development and testing environment debugging, educational demonstrations, etc. Summary: IQBandit fills the gap of product-grade management interfaces in the OpenClaw ecosystem, converting command-line and configuration file operations into an intuitive web interface, lowering the barrier to using self-hosted AI infrastructure, and is a worthy open-source solution to consider.