stickman/ROADMAP.md

Stickman Sandbox Builder - Project Roadmap

A kid-friendly (ages 10–15) interactive physics sandbox where users build environments, script stickman behaviors, and run live simulations.

---

1. Core Feature Breakdown

Environment & Physics Foundation

• Terrain Construction: Static environment pieces (floors, ramps, steps, walls) using StaticBody2D with grid snapping and auto-tiling collision boundaries.
• Dynamic Props: Rigid-body objects (boxes, balls, platforms) with customizable physical properties (gravity, mass, bounce, friction).
• Dynamic Ragdoll Engine: Seamless transition for stickmen between kinematic/animation-driven movement and dynamic RigidBody2D ragdoll physics upon impact, falling, or manual triggers.

Character Action & Directives

• Point-and-Click Navigation: Pathfinding and waypoint movement system for walking, running, and climbing to target positions.
• Action & Speech Queues: Sequenced action engine allowing stickmen to speak (speech/thought bubbles) and perform contextual animations (jump, wave, dance).
• IK Environment Interaction: Contextual IK targeting for gripping props, sitting on ledges, or stepping on uneven terrain.

Kid-Friendly UI & Canvas

• Palette Workbench: Icon-driven sidebar categorization (Terrain, Props, Actors, Logic) for drag-and-drop creation.
• In-World Context Menus: Popover radial menus on selected objects for quick actions (e.g., Set Speech, Assign Action, Edit Color, Toggle Ragdoll).
• Simplified Gizmos: Intuitive, clutter-free handles for scaling, rotating, and moving objects in 2D space.

Visual Logic & Event Triggers

• Event-Action Rules: Simplified cause-and-effect block system (e.g., When [Stickman] touches [Red Button] → [Drop Box]).
• Area Triggers: Placeable sensor zones (Area2D) that fire scene events when characters or props enter/exit.

Scene Management & State

• Play / Edit Mode Toggle: Instant mode switching—Edit Mode freezes physics for building; Play Mode unfreezes physics and runs interactive behaviors.
• Save/Load System: JSON serialization engine to preserve scene layouts, prop positions, customized stickmen, and event scripts.

---

2. Recommended Implementation Roadmap

Phase	Focus	Core Deliverables	Target Architecture
Phase 1	Physics & Ragdolls	Terrain shapes, prop physics, and Kinematic-to-Ragdoll swap	RigidBody2D + PinJoint2D
Phase 2	Play / Edit Engine	World building harness, object spawner, selection & transform gizmos	State Machine & Canvas Layer
Phase 3	Character Actions	Waypoint pathfinding, speech bubbles, animation/action queuing	NavigationAgent2D + Action Runner
Phase 3a (✅)	Core Director Functionality	Direct tool + action popup, per-stickman action queues, waypoint/badge visuals, nav mesh + runner	NavigationAgent2D + NavigationRegion2D + Action Runner
Phase 4	Visual Logic	Sensor zones, trigger-action mapping, interactive props	Event Bus & Logic Nodes
Phase 5	UX Polish & Saving	Radial context menus, sidebar palette, JSON scene serialization	UI Theme & JSON Parser

---

3. Detailed Phase Breakdown

Phase 1: Physics & Ragdoll Foundation

1. Build static terrain blocks (StaticBody2D) with slope and step collision.
2. Create dynamic prop types (RigidBody2D) with mass and surface material presets.
3. Implement ragdoll transition: detach visual bones/rig from animation drivers and bind to physics joint network on trigger.

Phase 2: Edit vs. Play Sandbox Harness

1. Implement central GameManager handling EDIT and PLAY states.
2. Create object spawner interface allowing drag-and-drop placement from UI into the world.
3. Add hover, selection outline, and 2D transform gizmos (translate, rotate, delete) active during EDIT mode.

Phase 3: Character Command System

1. Integrate NavigationAgent2D on stickmen for walking across user-built terrain. (✅ Phase 3a: code-built NavigationRegion2D re-baked on terrain edits; walk_to paths across placed blocks.)
2. Build a directive queue manager (WalkTo, PlayAnim, SayText, Wait). (✅ Phase 3a: walk_to / speak / wait / ragdoll / recover actions + per-rig action queue + queue runner.)
3. Add floating UI speech bubbles tied to the Head bone transform. (✅ Phase 3a: SpeechBubble, a world-space _draw() bubble child of the rig above the head.)

Phase 4: Logic & Triggers

1. Create TriggerArea2D nodes with visual boundaries visible in Edit Mode.
2. Build lightweight node-based or dropdown event connector (Trigger -> Target -> Action).
3. Implement interactive props like buttons, levers, and trapdoors.

Phase 5: UI Refinement & Persistence

1. Build radial popover context menu for selected objects.
2. Polish UI layout, visual feedback, and sound effects for kid accessibility.
3. Implement full scene save/load pipeline targeting .json format.