Thank you all for the kind words, I'm glad this is proving of some use. I find myself relying on it on a daily basis now.
@MinHyuk Lee From a security perspective I have done my best to ensure it is as secure as possible. This is what I have done:
Intercepted the Genesys's Architect SDK console output
Genesys's Architect SDK, which prints credentials to the console. I run the script using Architect SDK as a separate process and intercept all logs. This means I prevent the credentials getting to Claude Code.
OAuth Credentials for the plugin are managed by Claude's CLI, not my plugin
I intentionally leaned on Claude's own credential management for plugins so it can handle the storage of them in your local keychain.
Tooling is kept relevant to Architect
The plugin and tooling is designed specifically for Architect related activities. This means the scope of your OAuth Client's permissions can be kept small and focused only on Architect related tasks.
If anyone can think of other ways to improve the security of the plugin then I am always open to suggestions.
------------------------------
Lucas Woodward
Winner of Orchestrator of the Year, Developer (2025)
LinkedIn - https://www.linkedin.com/in/lucas-woodward-the-dev
Newsletter - https://makingchatbots.com
------------------------------
Original Message:
Sent: 09-26-2026 14:14
From: MinHyuk Lee
Subject: Create Architect Flows with AI (Claude Code)
When configuring Genesys Cloud using Claude, are there any security concerns we should take into account? Or are there things we should avoid?
------------------------------
MinHyuk
na
------------------------------