settled-reach/docs/workshops/control-interaction/workshop-outcomes.md

Workshop Outcomes: Control & Interaction Scheme

Workshop: Control & Interaction Scheme — "How Does the Player Touch the World?" Date: 2026-02-13 Rounds: 2 + final lead calls Participants: Gestalt, Ozzie, Tyre, Stig, Dudley, Paula, Araminta, Nigel Documenter: Qatux Full notes: docs/workshops/control-interaction/workshop-notes.md

---

Decisions Produced

This workshop produced 13 decisions ready for formalization. Suggested D-NNN IDs and domain file assignments below.

D-053: Movement as stance toggle system (scope.md)

• Decision: Movement uses a stance toggle ladder: Sprint / Walk / Careful / Crouch / (Prone future). Toggle-based, not hold-to-activate. Each archetype has a default stance via MovementProfile component. All characters have access to all stances. Prone: toggle out only in normal play; future combat allows "hit the deck" quick-entry.
• Rationale: Stances compose with perception (sprint = tunnel vision, careful = heightened awareness) and with the tile-based eavesdropping system. Toggle avoids modifier-key fatigue.
• Raised by: Lead (final call), building on Gestalt (triad), Dudley (MovementProfile), Nigel (character-defining speed)
• Server values (Dudley): Walk = 1 tile/2 ticks. Sprint = 1 tile/1 tick. Careful = 1 tile/3 ticks. Crouch TBD.
• Perception coupling (Gestalt + Ozzie, endorsed by lead): Sprint suppresses interpretation (monologue at 40%), not data (overlays still render). Careful enhances monologue (150%) and grants "tell notice bonus." Composable via multiplicative formula: movement_modifier * perception_load_modifier.
• Dissent: None after lead call. Nigel's character-differentiation concern addressed by MovementProfile (different defaults per archetype).

D-054: Tile-based movement with same-tile occupancy (architecture.md)

• Decision: All movement is tile-based (server-authoritative, discrete positions). Client-side Tween interpolation (100-150ms) hides the grid. Same-tile occupancy via TilePresence component (Standing/Prone/Seated/Fixture layers) allows multiple entities on one tile in different postures. Mouse facing is client-side float; server receives facing octant only.
• Rationale: Determinism (D-010). Tile-based enables shadowcasting, pathfinding, and trivial collision. Occupancy system addresses Nigel's emergent positioning needs (doorway blocking, eavesdrop positioning, furniture interaction) within tile-based constraints. ~150 lines server-side.
• Raised by: Tyre (tile-based, non-negotiable), Dudley (tiles-per-tick), Nigel (converted: "tiles are BETTER for replayability — discrete positions = finite meaningful choices")
• Dissent: None. Nigel initially proposed free movement (Round 1) but converted in Round 2 after seeing how tile-based spatial puzzles create replayability.

D-055: Sprint explicitly suppresses interaction buffer (architecture.md)

• Decision: When in Sprint stance, the server explicitly clears the interaction buffer. No interaction verbs are computed or sent to the client during sprint. Anomaly monologue survives sprint (Ozzie's "sprint double-take").
• Rationale: Mouse gymnastics to click during sprint = bad UX. Explicit suppression is cleaner and deterministic. "Sprint double-take" preserves the feel that the character is still aware even when the player can't interact.
• Raised by: Dudley (explicit suppression), Ozzie (anomaly survival)
• Dissent: Gestalt argued physics handles it naturally (player passes through interaction radius too fast to click). Lead ruled explicit suppression for clarity.

D-056: Cursor states — insert-styled geometric (perception.md)

• Decision: Four cursor states using the neural insert's geometric visual language (Araminta's spec):
  • Default: four thin inward-pointing ticks, bloom, white-blue #c8d0e0
  • Entity hover: ticks expand outward (150ms), corner brackets frame entity, shifts to D-033 relationship color, verb tooltip in insert styling
  • Object hover: ticks rotate 45 degrees to X-shape, muted grey #8b8ba0 (amber #e8c547 if flagged)
  • Weapon aim: hard transition, ticks extend, center gap widens, lines thicken 1px to 2px, warm white #f0e8d8, NO bloom. Entity in sights tints to D-033 color.
  • All transitions 150ms linear, z-layer 7, never changes by zone or narrative state (D-045).
  • Cursor changes on LOS, not just proximity (Stig). Click range: ~2 sim tiles (= 1m per D-066). Weapon-selected mode suppresses interaction prompts unless Shift held (Ozzie: "combat intent trumps social intent").
• Rationale: Diegetic — cursor is the insert's own interface element. Consistent with D-048 (insert overlay aesthetic). Entity in weapon sights tinting to D-033 creates moral friction (aiming at a friend should feel wrong).
• Raised by: Araminta (visual spec), Stig (UX rules), Ozzie (weapon suppression)
• Dissent: None.

D-057: Entity interaction — vertical list, insert-styled (perception.md)

• Decision: Entity interactions use a compact vertical list (not radial). 2-4 options max, anchored to entity position. Insert-styled with Araminta's aesthetic. New options (unlocked by knowledge changes) highlighted with gradient glow background. Radial menu reserved for world menu only.
• Rationale: Variable-length text options (e.g., confrontation lines in character voice) break radial spatial memory. List handles 1-4 options cleanly. New-item glow signals "something changed" without the UX hazard of geometry transforming under the cursor. Labels render on z-layer 6 (insert overlay) — diegetic test: labels disappear if insert is off.
• Raised by: Stig (vertical list structure + diegetic test), Araminta (insert aesthetic). Lead resolved in Stig's favor on structure, Araminta's on styling.
• Dissent: Araminta argued for spoke radial (geometry transformation signals qualitative change). Lead rejected — moving goalpost under cursor is bad UX during aiming.
• References: Disco Elysium (world-embedded indicators), Darkwood (minimal cursor), Rimworld (right-click context list).

D-058: World menu — radial, 4 spokes (perception.md)

• Decision: Right-click opens a radial world menu. Four spokes: Observe (eye icon), Insert (phone), Comms (signal), Wait (clock). Drag-release for power users, click-click for newcomers. Insert-styled: geometric lines, thin spokes with icons, nearly transparent. Renders on z-layer 6. v0.1: 2 spokes only (Observe + Insert).
• Rationale: Radial works for world menu because items are fixed categories that don't vary by knowledge state. Spatial memory builds quickly (~10 minutes). Scaling to 5-6 spokes is straightforward.
• Raised by: Stig (structure + implementation), Araminta (insert aesthetic)
• Dissent: None.

D-059: Fog — shader-based, five layers, knowledge-graph-driven (perception.md)

• Decision: Fog is a shader-driven system (not particles) with five distinct layers:
  1. Clear (vision cone): soft gradient edge 6-8 sim tiles (= 3-4 visual tiles per D-066), no hard line (Darkwood approach)
  2. Light fog (peripheral): desaturated 40-50%, brightness -30%, animated Perlin noise (8-10s cycle), entity D-033 colors visible but reduced
  3. Deep fog (explored): near-monochrome, ~10% zone temperature tint (bar=warm, hub=cool, corridor=neutral), more pronounced noise (15-20s cycle)
  4. Unexplored + maps app: geometric wireframe outlines #333340
  5. Unexplored, no maps: solid near-black #12141a
  • Fog is knowledge-graph-driven: same fog shows different information per character based on their KG. "Fog is not darkness — it's the absence of your attention."
  • Sound pings: 2-3 concentric expanding rings (sonar-style), insert white-blue. Loud = 3 rings bright fast, quiet = 1 ring faint slow. Fade over 1.5s.
  • Recognized entity in fog: D-033 color glow + faint identifying silhouette feature + 0.8s breathing pulse + half-tile position drift (approximate, not exact).
  • Unrecognized entity: neutral grey #555566 blob, no features.
  • Performance: <1ms/frame total. Vision cone = PointLight2D. Fog = darkness + noise shader on CanvasGroup (Layer 5).
  • Soft perception degradation (diegetic scan-line interference) at high overlay load, replacing hard mode cap.
• Rationale: Fog must be crucial in v0.1 (lead directive). KG-driven fog is the primary replayability mechanism — same corridor, different character, different information visible. Shader approach confirmed technically trivial by Tyre.
• Raised by: Araminta (visual spec), Tyre (technical validation), Nigel (replayability case), Gestalt (signal framework)
• Dissent: None.

D-060: Cognitive delay for fog recognition (perception.md)

• Decision: When the player recognizes a heard/sensed entity in fog, recognition is NOT instant. Single cognitive delay: 0.6s base, 0.3s when observe_anomaly triggers (urgent context). Values are tunable. Monologue fires DURING the delay ("Those footsteps... that's Kael's walk"), not after. Visual transition: grey blob to D-033 color + silhouette over ~0.3s within the cognitive delay window. Natural recognition (organic resolve) feels different from sensor recognition (digital snap with biometric ID).
• Rationale: Recognition should feel like a cognitive event, not a UI update. The delay creates a moment where the player's brain and the character's brain are working together. Context-sensitive urgency (0.3s for anomalies) prevents the delay from feeling sluggish in tense situations.
• Raised by: Ozzie (timing values + urgency split, adopted), Gestalt (longer values, not adopted but playtesting may adjust), Araminta (visual transition spec)
• Dissent: Gestalt proposed 0.8-1.2s (longer, more contemplative). Lead chose Ozzie's shorter values as starting point, tunable.

D-061: Dialogue box — bottom screen, max 20% height, no portraits (perception.md)

• Decision: Dialogue occupies the bottom of the screen, max 20% height, max-width (not percentage-based). Layout: NPC speech top, player response options below, left-aligned. Max 3 response options visible. NO portraits (NPC is on screen). Monologue floats ABOVE the dialogue box on z-layer 7 — spatial separation allows monologue to contradict dialogue visually. Walk-away via WASD, dialogue fades over 300ms, no close button.
• Rationale: Game world stays live above the dialogue box. The player sees the NPC's body language while talking — no portrait needed. Monologue above + dialogue below = the character can think one thing while saying another. Max 3 options + invisible locks = player never knows what they're missing.
• Raised by: Stig (UI spec), Lead (20% height constraint, max-width directive)
• Dissent: Stig initially proposed 25% height and 50% width centered. Lead constrained to 20% height and max-width.

D-062: Invisible locked dialogue options (content.md)

• Decision: Dialogue options the player hasn't unlocked are completely invisible. No grayed-out options. No lock icons. No hint that more options exist. Exception: NPC holding back is communicated via monologue ("She changed the subject. Fast."), not via locked UI elements.
• Rationale: Four reasons (Gestalt): (1) Asymmetry — you don't know what you don't know. (2) Dopamine — new options appearing on repeat visits IS the reward. (3) Anti-metagaming — no checklist to complete. (4) Confrontation surprise — Confront option appearing for the first time is a dramatic moment. Strongest consensus point of the entire workshop.
• Raised by: Gestalt (proposal + rationale), Nigel (emphatic reinforcement: "non-negotiable")
• Dissent: None. Unanimous across all 8 participants.

D-063: Confrontation — same box, different weight (content.md)

• Decision: Confrontation uses the same dialogue UI as casual conversation. No separate confrontation mode. Different weight communicated through: (1) italic text in character's internal voice for confrontation options (regular: "Shift schedule" / confrontation: "I saw you in corridor B-7"), (2) pre-delivery monologue beat (1-2s: "This changes things. No taking it back."), (3) world responds — NPC shifts to Tier 2 animation, entity D-033 color may fade, monologue frequency spikes, available topics narrow. Walk-away mid-confrontation contaminates social space (NPC routine shifts, KG records incompleteness). v0.1 staging: proximity check + audio dip + text styling. Camera tighten deferred.
• Rationale: Confrontation should feel heavy because of what you're saying, not because the UI changed. The weight comes from pacing, voice, and consequence.
• Explicitly NOT: Separate confrontation UI, timed responses, visible relationship meter, correct/incorrect approaches.
• Raised by: Paula (mechanism), Ozzie (physical staging), Stig (UI container)
• Dissent: None.

D-064: Walk-away — three-phase consequences (content.md)

• Decision: Walking away from dialogue (WASD during conversation) triggers three phases: (1) Immediate break — dialogue fades over 300ms, silence. (2) NPC reacts — animation shifts, may call after you, routine may change. (3) KG records — incompleteness is logged (confrontation initiated but not completed), queryable, affects future dialogue/monologue/NPC behavior. Walk-away consequences vary by NPC tolerance threshold per seed — no universal social rules to metagame.
• Rationale: Walking away is an action with meaning. Leaving mid-confrontation is different from leaving mid-smalltalk. The KG recording means the game remembers what you started.
• Raised by: Paula (three phases), Stig (WASD mechanic, 300ms fade), Ozzie (consequences), Nigel (tolerance per seed)
• Dissent: None.

D-065: Smuggler inventory — knowledge-primary with physical evidence (scope.md)

• Decision: Knowledge is the primary "inventory" for all characters. The smuggler additionally gets a minimal physical inventory: 3 specific items in v0.1 (manifest copy, corridor access token, personal comm log). Capacity per archetype: smuggler 3-4 slots, detective 2. Carried items are PRIVATE (info boundary — not visible to other entities unless revealed via search/scan/confrontation). UI: pocket icons bottom-right, 40x40px, no empty slots displayed. Verbs: Take, Place. Server: carried items are world entities with CarriedBy component.
• Evidence presentation differs by archetype (Paula): Detective sees case-file-style (structured, what/where/when/source/confidence, insert suggests links). Smuggler sees personal notebook (organized by person, informal voice, no contradiction flags). Same underlying KG, different presentation layer.
• Rationale: Smuggler's word doesn't carry institutional weight — they need tangible proof. Detective's word IS evidence (institutional authority), so they're mostly KG-only. Three items demonstrate the risk/reward concept and differentiate the smuggler's gameplay loop.
• Raised by: Lead (smuggler needs inventory), Paula (three items + presentation split), Gestalt (knowledge-primary framework), Tyre (minimal implementation), Dudley (server model + info boundary)
• Dissent: Tyre initially argued zero physical items in v0.1 (saves 3-4 sprints). Adapted with SmallVec<3> implementation.

---

Decisions Endorsed (Confirmed from Prior or This Workshop)

These were raised in the workshop and confirmed, but may overlap with or extend existing decisions:

Topic	Status	Reference
Signal/answer framework for information access	New principle — recommend formalization	Gestalt R1, Ozzie R1
Eavesdrop: passive base + positioning enhances quality	Extends D-018	Dudley R2 (ListeningFocus)
Two-phase verb computation (simulation max then observer filter)	Architecture pattern	Dudley R1
SP auto-pause for implant, overlay for MP	Extends D-009 implications	Tyre R1, Lead final
Implant organized by data type, not quest	UX principle	Ozzie R1
No highlights on implant data	UX principle	Ozzie R1
Soft perception degradation (diegetic), not hard cap	Extends D-017	Araminta R2, Lead LD-04
MovementProfile per archetype	Extends D-053	Dudley R2
Message history: design KG schema now, build display later	Scope constraint	Stig R2, Paula R2, Lead LD-09
Sprint double-take (anomaly monologue survives sprint)	Extends D-055	Ozzie R2

---

v0.1 Minimum Viable Interaction Model

Control Map

Input	Action	Stance/Context	Notes
WASD	Move	Current stance	Tile-based, Tween interpolated 100-150ms
Toggle key (TBD)	Cycle stances	—	Sprint / Walk / Careful / Crouch
Mouse movement	Facing direction	All	Client-side float, server gets octant
Left click	Default interaction	Non-weapon	Context: cursor state determines action
Left click	Fire	Weapon aim	Cursor = weapon reticle
Right click	World radial menu	All	v0.1: 2 spokes (Observe, Insert)
Scroll wheel	Zoom	All	Fog boundaries unchanged
WASD during dialogue	Walk away	Dialogue open	300ms fade, 3-phase consequences
Shift (with weapon)	Suppress interaction	Weapon selected	Hold to interact while armed

Stance Table

Stance	Speed	Monologue Rate	Noise	Interaction Buffer	v0.1
Sprint	1 tile/1 tick	40% (urgent only)	Loud	CLEARED (D-055)	Yes
Walk	1 tile/2 ticks	100% (normal)	Normal	Active	Yes
Careful	1 tile/3 ticks	150% + tell notice bonus	Quiet	Active (enhanced)	Yes
Crouch	TBD	TBD	Very quiet	Active	Yes (D-054 occupancy)
Prone	TBD	TBD	Minimal	TBD	No (future)

Cursor States (D-056)

State	Visual	Color	z-layer
Default	Four thin inward ticks, bloom	#c8d0e0	7
Entity hover	Expand outward, corner brackets, verb tooltip	D-033 color	7
Object hover	Rotate 45 degrees (X-shape)	#8b8ba0 / #e8c547	7
Weapon aim	Extend, gap widens, 2px, NO bloom	#f0e8d8	7

Fog Layers (D-059)

Layer	Treatment	Key Values
Clear	Soft gradient edge	6-8 sim tiles (= 3-4 visual tiles, D-066)
Light fog	Desaturated, Perlin noise	40-50%, 8-10s cycle
Deep fog	Near-monochrome, zone temperature	~10% tint, 15-20s cycle
Unexplored + maps	Geometric wireframe	#333340
Unexplored	Solid black	#12141a

Entity Interaction (D-057)

Element	Spec
Menu type	Vertical list, insert-styled
Max options	2-4, anchored to entity position
New option highlight	Gradient glow background
Trigger	LOS + ~2 sim tiles proximity (= 1m, D-066)
Diegetic test	Labels on z-layer 6, disappear if insert off

World Menu (D-058)

Element	Spec
Menu type	Radial, 4 spokes (v0.1: 2)
v0.1 spokes	Observe (eye), Insert (phone)
Full spokes	+ Comms (signal), Wait (clock)
Interaction	Drag-release or click-click

Dialogue (D-061, D-062, D-063, D-064)

Element	Spec
Position	Bottom screen, max 20% height
Width	Max-width (not percentage) — value TBD
Layout	NPC speech top, options below, left-aligned
Max options	3 visible
Locked options	Invisible (D-062)
Portraits	None
Monologue	Floats above dialogue (z-layer 7)
Confrontation	Same box, italic voice, 1-2s monologue beat
Walk-away	WASD, 300ms fade, 3-phase consequences

Smuggler Inventory (D-065)

Element	Spec
Model	Knowledge-primary + physical evidence
Capacity	Smuggler 3-4, Detective 2
v0.1 items	Manifest copy, corridor access token, personal comm log
UI	Pocket icons, bottom-right, 40x40px, no empty slots
Verbs	Take, Place
Info boundary	Carried items private

Fog Recognition (D-060)

Element	Spec
Cognitive delay	0.6s base, 0.3s urgent
Monologue timing	Fires DURING delay, not after
Visual transition	Grey to D-033 color + silhouette over ~0.3s
Natural recognition	Organic resolve
Sensor recognition	Digital snap with biometric ID
Position accuracy	Half-tile drift (approximate)

Server Performance Budget

System	Budget	Source
Full tick pipeline (9 steps)	<0.5ms/tick	Dudley
Fog rendering	<1ms/frame	Tyre
Client overlay composition (4 modes)	~7-11ms/frame	Tyre

---

Implementation Priorities

Priority 1: Core Movement and Interaction (Server)

Estimate: ~5-6 days server (Tyre)

1. Stance system (Sprint/Walk/Careful/Crouch) with tick-based movement values
2. MovementProfile component per archetype
3. Sprint interaction buffer suppression (D-055)
4. Same-tile occupancy (TilePresence with posture layers) (~150 lines)
5. Two-phase verb computation extension (ObjectType component, KG-gated Phase 2)
6. Cognitive delay system for fog recognition (0.6s/0.3s, tunable)

Priority 2: Core UI and Rendering (Client)

Estimate: ~8-10 days client (Tyre)

1. Cursor state machine (4 states, Araminta spec, 150ms transitions)
2. Fog shader (5-layer Araminta spec, CanvasGroup Layer 5)
3. Fog entity visualization (sound pings, recognized/unrecognized, cognitive delay animation)
4. Entity interaction vertical list (insert-styled, z-layer 6, diegetic test)
5. World radial menu (2 spokes: Observe, Insert)
6. Dialogue box (max 20% height, max-width, NPC speech + options + monologue)
7. Walk-away mechanic (WASD detection, 300ms fade)

Priority 3: Smuggler Systems

1. Minimal inventory (SmallVec or BTreeMap, Take/Place, info boundary)
2. Pocket-icon UI (bottom-right, 40x40px, no empty slots)
3. Three smuggler items (manifest copy, access token, comm log) — content definition
4. Contraband detection (NPC scan checks carried items + KG)

Priority 4: Perception and Audio

1. Eavesdrop positioning (ListeningFocus accumulates stationary_ticks) — deferrable per Dudley
2. Sprint double-take (anomaly monologue survival)
3. 6 new audio assets (cursor hover, weapon mode, fog recognition, implant open/close, dialogue emerge, confrontation dip)
4. Confrontation staging (proximity + audio dip + text styling)

Priority 5: Deferred (Design Now, Build Later)

1. InteractionMemory KG schema for message history (LD-09)
2. Archetype-specific evidence presentation (detective case file vs smuggler notebook)
3. Pre-dialogue proximity acknowledgment (NPC turns, ambient line)
4. Camera tighten for confrontation staging
5. Prone stance
6. Perception mode soft degradation (scan-line interference)

---

Open Questions Deferred to Implementation

These questions were not resolved in the workshop and should be addressed during implementation:

#	Question	Owner Suggestion
OQ-01	Tile size in world units? Resolved: 0.5m sim tiles, 1m visual tiles (D-066)	Tyre/Dudley
OQ-07	No insert = no interaction prompts? (Diegetic test implication)	Stig/Gestalt
OQ-08	False-positive fog shapes: gameplay or atmospheric?	Gestalt/Araminta
OQ-09	Zone temperature memory: server-tracked or client-only?	Tyre/Dudley
OQ-10	Sensor recognition "different monologue voice": text styling or narrator tone?	Paula/Stig
OQ-13	No data highlights in implant — how does player know new data was added?	Ozzie/Stig
OQ-14	Scan effectiveness varies per character seed: different data or different actions?	Gestalt/Dudley
OQ-18	Nigel's dialogue access tiers vs D-041 confidence hierarchy: same or layered?	Paula/Gestalt
OQ-21	Archetype evidence presentation: skin on same app or different apps?	Stig/Paula
OQ-24	Inventory capacity: 3 slots (Tyre) or 4 (Dudley)?	Lead call needed
OQ-25	Stance name: "Careful" or "Sneak"?	Lead call needed
OQ-26	Multiplicative perception formula: actual modifier values?	Gestalt/Dudley
OQ-27	Soft perception degradation threshold?	Araminta/Gestalt
OQ-28	InteractionMemory KG schema design?	Paula/Dudley
OQ-29	Dialogue max-width: pixel value?	Stig/Lead
OQ-30	14 audio assets: who creates them, when in sprint schedule?	Inigo/Si

---

Suggested Tickets

For SI to process and assign to sprints:

Server Team

#	Title	Scope	Estimate	Depends On
T-1	Implement stance system (Sprint/Walk/Careful/Crouch)	D-053	2-3d	—
T-2	MovementProfile component per archetype	D-053	1d	T-1
T-3	Sprint interaction buffer suppression	D-055	0.5d	T-1
T-4	Same-tile occupancy (TilePresence layers)	D-054	1-2d	—
T-5	ObjectType component + verb sets per type	D-057 (server)	1d	—
T-6	Two-phase verb computation: KG-gated Phase 2	D-057 (server)	1-2d	T-5
T-7	Cognitive delay system for fog recognition	D-060	1d	—
T-8	Smuggler inventory (BTreeMap/SmallVec, Take/Place, info boundary)	D-065	1-2d	—
T-9	Contraband detection (scan + KG check)	D-065	0.5d	T-8
T-10	ListeningFocus (eavesdrop positioning)	D-053 ext	1d	T-1 (deferrable)
T-11	Walk-away KG recording (confrontation incompleteness)	D-064	0.5d	—
T-12	Sprint anomaly monologue (double-take)	D-055 ext	0.5d	T-3

Client Team

#	Title	Scope	Estimate	Depends On
T-13	Cursor state machine (4 states, Araminta spec)	D-056	2d	—
T-14	Fog shader (5-layer, CanvasGroup Layer 5)	D-059	3-4d	—
T-15	Fog entity visualization (pings, recognized/unrecognized, delay anim)	D-059, D-060	2d	T-14, T-7
T-16	Entity interaction vertical list (insert-styled)	D-057	1-2d	T-13
T-17	World radial menu (2 spokes v0.1)	D-058	1d	—
T-18	Dialogue box (max 20% height, max-width)	D-061	2d	—
T-19	Dialogue response selection (max 3, invisible locks)	D-062	1d	T-18
T-20	Confrontation text styling (italic voice, monologue beat)	D-063	1d	T-18, T-19
T-21	Walk-away mechanic (WASD detect, 300ms fade)	D-064	0.5d	T-18
T-22	Smuggler pocket-icon inventory UI	D-065	1d	T-8
T-23	Stance toggle UI (keybind, HUD indicator)	D-053	0.5d	T-1

Audio Team

#	Title	Scope	Estimate	Depends On
T-24	6 interaction audio assets	D-059, D-060, D-061	2-3d	—

Design/Content Team

#	Title	Scope	Estimate	Depends On
T-25	Define 3 smuggler item specs (content/attributes/interactions)	D-065	1d	—
T-26	InteractionMemory KG schema design	D-064, LD-09	1d	—
T-27	Archetype evidence presentation spec (detective case file vs smuggler notebook)	D-065	1d	—

---

Replayability Validation

Nigel's Round 2 scenario trace validates the v0.1 interaction model against D-027 replayability criteria:

• 10-minute scenario, two characters (smuggler vs detective), same world state
• 12 points of divergence, zero scripted branching
• All divergence emerges from knowledge graph differences interacting with: tile-based positioning, fog visibility, verb sets, dialogue access tiers, evidence presentation, and inventory constraints
• Passes the second-playthrough test: playing again as a different character reveals genuinely different information, interactions, and available actions

---

Cross-Layer Interaction Chains (Workshop Validated)

Five chains documented in workshop notes demonstrate how the layers compose:

1. Eavesdropping Loop: Careful stance > tile positioning > ListeningFocus > monologue > KG update > new dialogue options
2. Fog Recognition Loop: Sound/sensor detection > cognitive delay (0.6/0.3s) > monologue during delay > KG query > identity resolution or blob remains
3. Confrontation Chain: KG contradiction > monologue signal > approach NPC > pre-dialogue beat > dialogue > confront option (italic) > monologue beat > delivery > world responds > KG records
4. Smuggler Evidence Loop: Eavesdrop > KG records discrepancy > terminal access > raw data (no highlights) > player notices > Take physical item > info boundary (private) > risk/leverage
5. Sprint Double-Take: Sprint through corridor > interaction suppressed > routine monologue suppressed > anomaly detected > delayed monologue > player decides: double back or continue

---

Workshop closed. 13 decisions produced (D-053 through D-065). 27 suggested tickets across server, client, audio, and design teams. Full discussion notes at docs/workshops/control-interaction/workshop-notes.md.

Compiled by Qatux.