Royale Morale is an open-source AI agent designed to help governments and organizations monitor and evaluate Agentic AI systems. It provides automated way to ensures these systems adhere to ethical standards and safety protocols, providing oversight and maintaining public trust in AI technologies. Royale Morale is under heavy development and encouraging contributions from the community.
Royale Morale focuses on building an open-source agent that scans source code and analyzes AI prompts used in developing Large Language Model (LLM) applications. The primary aim is to ensure AI Safety and AI Ethics in these applications.
This project began back in 2016 with the goal of creating an easy way to build AI ChatBots and Agents through IBM Watson's AI. But now evolving to become a Safety Agent for the Agentic world of AI. The contributors on this repository have experience in AI and also in helping governments in their National AI Strategies.
At the moment this is focussed on agents and LLM applications that are built on Ruby & Rails. It scans the source code automatically, detects for LLM related code and AI Prompts and uses a configuration files to run a AI Athics & Safety Ranking.
-
Source Code Scanning: Automatically scans and parses source code to identify AI prompts.
-
AI Prompt Analysis: Evaluates AI prompts for compliance with ethical standards and safety protocols.
-
Benchmarking: Uses advanced models to benchmark AI prompts against recognized AI ethics principles.
(easy way to use it coming soon in next few days)
Highly encourage contributions from the community.
This version of Royale Morale is released under the MIT License.
Let @parolkar know.
- ForwardLeague.org - Community of Forward Thinkers
- Dotted Circle