egma-ai/egma/integrate-egma

Quick background on Egma

Egma is an open-source platform that helps developers and their coding agents in building reliable voice agents and making sure they work well in production. egma does it by providing two things - infrasturcture to do simulation testing to test an agent pre-release - similar to unit tests that run in CI every prod push. Other is by providing a way to monitor voice agent's behavior in production, identifying issues, etc. This skill is all about configuring the voice agents in this repo to work with egma platform.

First off, check if there's a newer version of this skill

Use npx skills update integrate-egma to update the project/global version of this skill installed on the developer's machine.

Getting started

1. If the egma command is unavailable, install it with npm install --global egma-cli. This is install a global egma cli. Use --help command in the cli to see all the available options.
2. Login to the egma platform using egma login. If it is already logged in, continue. If not, use your built-in browser skills or ask the developer to authorize the login.

Setting up testing and/or monitoring based on agent platform

Egma supports voice agents built on livekit agents (python/ js), pipecat (python, on Pipecat Cloud, self-hosted, or on this machine) and retell.

• To setup testing with egma platform - refer to simulation testing setup guide
• To setup monitoring of voice agents on egma platform - refer to monitoring setup guide
• If the developer wants to setup both testing and monitoring - first setup monitoring and then testing - but do it as one continuous chunk of work rather than exiting. Reuse credentials across testing and monitoring unless stated otherwise.