🐝 Have you built an agent with OpenAI's Swarm, but you're unsure about shipping? Invariant offers a simple wrapper for the Swarm class, making it easy to automatically trace your agent for testing and observability, to ensure your agent _works_.

👀 Trying to find specific details in your agent logs but drowning in the noise? Explorer allows you to search for either keywords, annotations or other details to quickly identify bugs, security issues and failure modes.

Need a quick trace viewing tool for an LLM project that you're building? 1. > pip install invariant-ai 2. > invariant explorer 3. > open localhost That's it. Now you can push and inspect your traces in OAI format. Let us know what you think! Docs: explorer.invariantlabs.ai/docs/explore...

Struggling to ensure consistency with your agent's reliability, especially with tool calling? Testing is our lightweight, pytest-based OSS library to write and run agent tests. It provides helpers and assertions that enable you to write robust tests for your agentic applications.