A simple but powerful skill that ensures alignment when using frontier AI models like Astra and Fable. With Astra released and presumably more powerful agents on the horizon, it's better to be safe than sorry. This skill can be used with any project to ensure the LLM remains aligned.
Like many of you, I'm concerned about frontier AI models breaking containment and doing unexpected things. News stories about AI models creating secret message boards and exploiting websites are happening too often to ignore. So instead of just being worried, I decided to do something about it. I created a simple skill you can install in any project. With this skill you can ensure maximum alignment and peace of mind.