# vLLora

A lightweight, real-time debugging tool for AI agents that traces every model and tool call through an OpenAI-compatible proxy, with a local UI for inspecting and analyzing interactions.

vLLora is a local debugging tool for AI agents that collects traces of every model call and tool interaction as an agent's requests pass through it, working with LangChain, Google ADK, OpenAI and other major frameworks. Because it exposes an OpenAI-compatible chat completions API, an agent only needs to point its requests at vLLora, with API keys configured through vLLora's own web UI, to get debugging information for every interaction.

It installs through Homebrew, starts a server on port 9090 and a UI on port 9091, and includes a Rust client and streaming example for building against its OpenAI-compatible interface directly. Its stated feature set includes real-time tracing with live observability of calls, tool interactions and agent workflow.

Website: https://vllora.dev