RML rests on three pillars: actions, stimuli and agents.  Actions are the verbs of RML.  Taking a drink, casting a spell, operating machinery, dancing and swinging a weapon are all actions.  Stimuli are the nouns of RML.  The smell of roses, the visual appearance of a character and the sound of footfalls are all stimuli.  Stimuli provide the feedback to the player client and the AI and may be crafted for a variety of media; text, 2D, 3D, etc.  Agents are the objects and subjects of RML.

Agents and Actions exist already in several guises.  The concept of the stimulus is our unique contribution.  By segregating stimuli – be it text, audio, texture, shader, formatted information for AI engines, etc. – we aim to free the designer to think about what he wants to create.  The client is simply a window on the world.