Utility-HTN Manager

Adds a simple stimulus event. Use case: raise alert when a guard hears a noise.

Adds a detailed stimulus from JSON. Use case: set custom decay and update world state in one call.

Removes current stored stimuli. Use case: reset sensory pressure after a scripted state change.

Lowers alert by a fixed amount. Use case: calm an agent after the player escapes.

Sets alert level directly from 0 to 1. Use case: force combat mode when player is spotted.

Adds or ensures a compound task node in the current builder draft.

Creates a method branch on a compound task.

Adds one world-state condition gate to a method.

Adds one subtask to a method in execution order.

Adds a primitive task that resolves to one runtime task id.

Adds one scorer input with a 2-point curve.

Starts a clean task-network draft for one agent type.

Starts a clean utility-scorer draft.

Removes the in-progress task-network draft for one agent type.

Removes the in-progress utility-scorer draft.

Finalizes and registers the current task-network draft.

Finalizes and registers the current utility-scorer draft.

Links a utility scorer to one method branch.

Puts one agent in a squad. Use case: build patrol teams at layout start.

Finds nearest free slot and reserves it. Use case: auto-pick closest cover for each squad member.

Removes one shared squad key. Use case: clear temporary rally point after regroup.

Marks squad members stale for next planning update. Use case: tactic changed for full squad.

Loads many slot positions from JSON array data. Use case: import a pre-authored cover map.

Loads slot positions from a world-state key. Use case: share dynamic slot maps through state data.

Releases a reserved slot. Use case: free cover slot when an agent moves away.

Removes one agent from its squad. Use case: cleanup when an NPC is defeated.

Requests immediate planning for squad members. Use case: execute synced reaction now.

Reserves a squad slot with TTL. Use case: allocate unique cover nodes to avoid overlap.

Defines one slot position from event-sheet coordinates. Use case: set cover node positions at layout start.

Combined reserve/release slot action. Use case: simplify event sheets while keeping full slot control.

Sets the leader agent for a squad. Use case: only leader chooses squad tactic.

Writes shared squad state. Use case: set tactic to flank_left for all members.

Combined invalidate/request squad planning action. Use case: choose delayed or immediate squad replan with one ACE.

Checks current squad membership. Use case: run squad-only logic per agent.

Checks if a slot is available. Use case: only reserve cover if not already occupied.

Checks leader status in current squad. Use case: leader-only tactic writes.

Triggered when a slot is released or expires. Use case: make slot visible as available again.

Triggered when a slot gets reserved. Use case: update debug overlays for cover claims.

Triggered when an agent joins or leaves a squad. Use case: refresh squad HUD widgets.

Triggered when shared squad key-values change. Use case: react when tactic flips.

Gets the squad ID for an agent. Use case: print squad tags in debug text.

Gets assigned slot ID for an agent and slot type. Use case: drive movement to assigned cover node.

Counts members in a squad. Use case: detect when reinforcements are needed.

Counts active squads. Use case: monitor coordination load.

Gets squad member UID at index. Use case: iterate members in a repeat loop.

Gets owner UID of a slot. Use case: verify which agent owns a cover slot.

Gets leader UID for a squad. Use case: route leader-only commands.

Reads one shared squad state key. Use case: followers read current tactic.

Prints one agent snapshot in the console. Use case: quickly check world state and current plan.

Turns debug logging on or off. Use case: inspect planner decisions while testing.

Runs one planning pass immediately. Use case: apply state changes now instead of waiting.

Turns processing on or off. Use case: disable AI while game menus are open.

Caps agents evaluated per update. Use case: smooth frame-time spikes with large NPC groups.

Sets time between planning passes in seconds for Interval mode. Beginner tip: try 0.1 first, then raise to 0.2 or 0.25 if you need more performance.

Switches how often planning runs. Beginner tip: use Per frame while learning, then switch to Interval if frame rate drops with many agents.

Sets max planning time allowed per update in seconds. Beginner tip: start at 0 (off), then try 0.001-0.003 if big AI spikes cause frame drops.

Removes remaining planned tasks. Use case: keep an agent idle until new orders arrive.

Forces one task to run first. Use case: trigger a scripted reaction during a cutscene.

Marks a plan as outdated for next update. Use case: world changed but current task should finish.

Stops planning updates for one agent. Use case: freeze NPC AI while a dialogue plays.

Builds a fresh plan right now. Use case: replan instantly after a major event.

Restarts planning for one agent. Use case: resume AI after a stun or cutscene.

Removes one world-state key from an agent. Use case: clear last known target after search ends.

Loads a task tree from JSON for one agent type. Use case: register guard AI at layout start.

Loads a utility scoring rule from JSON. Use case: tune attack vs retreat behavior by data.

Sets a value shared by all agents. Use case: toggle a global alarm state.

Sets one value for one agent. Use case: update health or target visibility each tick.

True if a world-state key exists. Use case: guard logic when optional data may be missing.

True when planning is paused for that agent. Use case: skip AI updates during scripted moments.

True when an agent has work to do. Use case: branch events when an NPC is currently busy.

Checks the current alert tier name. Use case: run different events for suspicious vs combat.

True when a task network is loaded for that type. Use case: verify setup before spawning agents.

True when a scorer ID is available. Use case: validate data packs before combat starts.

Triggered when alert tier changes. Use case: change music or UI for suspicion/combat states.

Triggered when no valid plan is found. Use case: force an idle or safe fallback action.

Triggered when high-urgency input interrupts a plan. Use case: swap from patrol to combat immediately.

Triggered when a primitive task begins. Use case: play movement or attack events for that task ID.

Triggered when any stimulus is added. Use case: spawn investigation markers at stimulus coordinates.

Triggered after task completion is reported. Use case: chain to the next animation state.

Triggered when a task fails. Use case: run a fallback task or recovery behavior.

Triggered after a network loads. Use case: print setup success in debug events.

Current active task for one agent. Use case: show simple behavior labels above NPCs.

Alert level value from trigger context. Use case: drive bar fill in a stealth meter.

Numeric alert tier from trigger context. Use case: simple tier-based UI indicators.

Named alert tier from trigger context. Use case: show readable debug text in HUD.

Completed task ID from trigger context. Use case: reward points for finished objective tasks.

Current started task ID. Use case: branch events by task name like patrol or chase.

Returns scorer value for an agent now. Use case: graph utility output while balancing AI.

Compound task where planning failed. Use case: log broken network branches while testing.

Failed task ID from trigger context. Use case: retry or switch to a safer fallback action.

Reads one shared global-state value. Use case: branch logic when global alarm is active.

Task that got interrupted. Use case: stop old animations cleanly before new behavior starts.

Score that caused the last interruption. Use case: tune threshold to avoid over-reacting.

New task after interruption. Use case: immediately start VFX for emergency reactions.

Remaining task count in the plan. Use case: detect when agents are close to idle.

Task ID at a plan index. Use case: preview next action for debugging UI.

Previous task ID before current one. Use case: detect when behavior transitions happen.

Previous alert tier index. Use case: detect if alert increased or decreased.

Last registered agent type key. Use case: verify setup order in startup events.

Stimulus intensity value. Use case: scale reaction speed by how strong the event was.

Stimulus type string from trigger context. Use case: react differently to sound vs visual cues.

Stimulus X coordinate. Use case: send NPC to investigate that position.

Stimulus Y coordinate. Use case: mark a noise location on minimap.

Agent UID from the latest trigger. Use case: pick the exact instance to move or animate.

Reads one world-state value for an agent. Use case: display health or awareness in UI.

Current registered agent count. Use case: monitor AI population for performance tuning.

How many task networks are loaded. Use case: quick startup validation in debug builds.

How many utility scorers are loaded. Use case: confirm balance data loaded correctly.

Number of primitive tasks for one type. Use case: verify network content during setup.

Gets agent UID by list index. Use case: loop through all managed agents.

Gets primitive task ID by index. Use case: print all registered tasks to a debug panel.