A Goal is a concrete, actionable intention formed by an agent to resolve an internal tension or respond to an external stimulus. It is the bridge between the abstract landscape of Need and Belief, and the practical world of planning and action.

Within ATET, a Goal is not a “mission objective” handed down by the simulation. It is an emergent property of the agent’s own mind. It is the conscious articulation of a desire, the first step an Incarnation takes in attempting to impose its will upon the Tapestry. Where a Need is a feeling, a Goal is a purpose.


The Nature of Goals

Goals are the engine of agent behavior. They provide the “what” that drives the Plan and Appraise phases of the cognitive loop. A hungry agent doesn’t just feel hunger; it forms the SeekSustenanceGoal. A threatened agent doesn’t just feel fear; it forms the SeekSafetyGoal.

This formalization is critical. It transforms a diffuse, internal state into a specific, addressable problem that the agent’s logic can work to solve. The simulation is a constant churn of agents generating, prioritizing, pursuing, and resolving these emergent Goals.


How Goals Arise

An agent’s mind is a crucible where the pressures of its existence are forged into intent. Goals are generated by the GoalGenerationSystem in response to several key triggers:

  • From Unmet Need: This is the most common and powerful source. As detailed in the Agent Cognitive & Behavioral Systems, when a Need component’s value crosses its goal_trigger_threshold, a corresponding Goal is generated. A critical SustenanceNeed directly creates a SeekSustenanceGoal. This is the core of the survival simulation.

  • From Perceived Opportunity or Threat: The PerceptionSystem provides a constant stream of information. A Belief can frame this information as an opportunity or a threat, triggering a Goal.

    • Opportunity: An agent with a Belief that “Rare Flora are valuable” perceives a rare flower. This can generate an AcquireItemGoal.
    • Threat: An agent with a Belief that “Wolves are dangerous” perceives a wolf. This immediately generates a SeekSafetyGoal, often with a very high priority.
  • From Quest Participation: A player or NPC engaged in a Quest will generate Goals related to its resolution. A quest to “uncover the truth of the ruin” will spawn specific sub-goals like InvestigateLocation(Ruin) or Translate(AncientInscription).

  • From Social Interaction: A conversation can implant a Goal. Being persuaded by another agent might create a HelpAgentGoal. Being insulted might create a RetaliateSociallyGoal.


The Life of a Goal

A Goal is a temporary component attached to an agent, containing crucial data for the AI’s decision-making processes.

  • Priority: Every Goal has a calculated priority. This is not a static value. It is dynamically assessed by the GoalPrioritizationSystem, which weighs the urgency of the underlying Need, the immediacy of the perceived threat, and the agent’s own Beliefs about what is important. An agent may have both a SeekSustenanceGoal and a SeekSafetyGoal, but the GoalPrioritizationSystem will elevate the safety goal to the CurrentActiveGoalFocus if a predator is two feet away.

  • Pursuit and Planning: Once a Goal becomes the CurrentActiveGoalFocus, the PotentialActionDerivationSystem and ActionAppraisalSystem work to create and evaluate plans to achieve it. A SeekSustenanceGoal might generate potential actions like [MoveTo(BerryBush), Eat(Berries)] or [MoveTo(Town), TradeFor(Food)].

  • Resolution: A Goal is resolved when its completion condition is met. The GoalResolutionSystem monitors for these conditions.

    • A SeekSustenanceGoal is resolved when the SustenanceNeed rises above its trigger threshold.
    • A SeekSafetyGoal is resolved when the perceived threat is gone.
    • An AcquireItemGoal is resolved when the item is in the agent’s inventory. Upon resolution, the Goal component is removed from the agent, freeing its cognitive resources to focus on the next most important Goal.

The Player’s Experience

The player is not typically aware of the formal “Goal” components driving their Incarnation or the NPCs around them. Instead, they experience this system through its effects on the Subjective Interface and emergent behavior.

  • Internal Monologue: When your Incarnation forms a high-priority Goal, it might manifest as a line of internal monologue text: “I’m starving. I need to find something to eat, now.” This is the narrative expression of the SeekSustenanceGoal becoming the CurrentActiveGoalFocus.
  • Observing Intent: The player learns to read the Goals of others through their actions. An NPC suddenly walking purposefully towards the river is likely driven by a SeekWaterGoal. An NPC who keeps glancing nervously at the player while backing away is clearly acting on a SeekSafetyGoal.
  • Creating Goals: The player’s primary method of agency is in creating Goals for themselves. Deciding to explore a cave, talk to a specific person, or build a shelter are all acts of conscious Goal creation that direct their Incarnation’s subsequent actions.

Design Priorities

  • Emergent Motivation: Goals should arise organically from the agent’s internal state and its interaction with the world, providing a constant, believable stream of motivation.
  • Grounded in Simulation: Every Goal should be traceable back to a concrete driver—a Need, a Belief, a perception, or a social commitment. This keeps agent behavior from feeling arbitrary.
  • Dynamic and Rational Prioritization: The Goal system is the core of the agent’s apparent rationality. Its ability to dynamically prioritize between competing desires (like hunger and safety) is fundamental to creating believable and intelligent behavior.