- Added hand frames PNG for interaction modes.
- Updated HUD CSS to support new player HUD buttons and styles.
- Enhanced combat configuration to define interaction modes: interact, jab, and cross.
- Integrated player HUD creation and management in the game core.
- Improved player combat system to handle interaction modes and associated animations.
- Modified interaction handling to auto-switch modes for chairs and hostile NPCs.
- Updated health UI to always display health status.
- Created a new HUD JavaScript module to manage avatar and interaction mode toggle.
- Added a test HTML file for the three-mode toggle functionality.
- Deleted unused character images: woman_in_science_lab_coat.png and woman_with_black_long_hair_bow_in_hair_long_sleeve_(1).png.
- Added new padlock icon asset for UI.
- Introduced player_preferences.css for styling the player preferences configuration screen.
- Updated game.js to load new character atlases with simplified filenames.
- Enhanced player.js to create custom idle animations for characters.
- Implemented sprite-grid.js for sprite selection UI, including a preview feature.
- Updated database schema to include break_escape_player_preferences table for storing player settings.
- Modified convert_pixellab_to_spritesheet.py to map character names to simplified filenames and extract headshots from character images.
- Updated the game to support new character sprite atlases for both male and female characters, allowing for a wider variety of NPC designs.
- Improved player sprite initialization to dynamically select between atlas-based and legacy sprites, enhancing flexibility in character representation.
- Refined collision box settings based on sprite type, ensuring accurate physics interactions for both atlas (80x80) and legacy (64x64) sprites.
- Enhanced NPC behavior to utilize atlas animations, allowing for more fluid and diverse animations based on available frames.
Files modified:
- game.js: Added new character atlases and updated sprite loading logic.
- player.js: Improved player sprite handling and collision box adjustments.
- npc-behavior.js: Updated animation handling for NPCs to support atlas-based animations.
- npc-sprites.js: Enhanced NPC sprite creation to accommodate atlas detection and initial frame selection.
- scenario.json.erb: Updated player and NPC configurations to utilize new sprite sheets and animation settings.
- m01_npc_sarah.ink: Revised dialogue options to include new interactions related to NPCs.
- Updated the games_controller to remove starting items from the inventory response, ensuring no duplicates are sent to the client.
- Enhanced the inventory processing logic in JavaScript to check for existing items by ID, type, or name before adding new items, preventing duplicate entries in the player's inventory.
Files modified:
- games_controller.rb: Adjusted inventory response handling.
- inventory.js: Added duplicate item check during inventory processing.
- Implemented keyboard event handling to allow players to use the spacebar to continue and number keys (1-9) to select choices in the minigame.
- Updated UI to display choice buttons with number prefixes for easier selection.
Files modified:
- person-chat-minigame.js: Added keyboard event listener for choice selection.
- person-chat-ui.js: Updated choice button creation to include number prefixes.
- Added a new task to report the discovery of Operation Shatter to SAFETYNET, including dialogue for Agent 0x99.
- Introduced new variables to track player interactions regarding the operation and its implications.
- Updated scenario files to reflect the urgency and details of the operation, emphasizing the calculated casualties and the psychological warfare aspect.
- Shortened the opening briefing to streamline the introduction of the mission and its stakes.
Files modified:
- scenario.json.erb: Added new task for reporting Operation Shatter.
- m01_closing_debrief.json: Updated dialogue to reflect new mission details.
- m01_derek_confrontation.json: Adjusted confrontation dialogue to align with new narrative elements.
- m01_npc_maya.ink: Enhanced Maya's dialogue regarding Operation Shatter.
- m01_opening_briefing.ink: Streamlined introduction and urgency explanation.
- m01_phone_agent0x99.ink: Added reporting mechanics for Operation Shatter discovery.
- Changed CSS file paths in index.html and show.html.erb to reflect new directory structure under public/break_escape.
- Added a new tutorial.css file with styles tailored for the tutorial system, ensuring a cohesive pixel-art aesthetic.
- Enhanced the tutorial manager to track player interactions, including clicks to move and inventory item usage, improving the tutorial experience for new players.
- Updated tutorial steps to dynamically include objectives based on the current scenario.
Files modified:
- index.html: Updated CSS links.
- show.html.erb: Updated CSS links.
- tutorial.css: New styles for the tutorial system.
- player.js: Added notifications for player actions.
- tutorial-manager.js: Enhanced logic for tracking tutorial progress and objectives.
- interactions.js: Added notifications for inventory interactions.
- Refactored the MissionsController to use local variable `missions` instead of instance variable `@missions` for better clarity and performance.
- Introduced grouping of missions by collection for display, improving the organization of mission listings.
- Updated the view to iterate over `@missions_by_collection`, enhancing the presentation of missions by their respective collections.
- Added new CSS styles for collection headings to improve UI aesthetics.
Additionally, removed invalid missions from the database and updated mission collections for better categorization.
Files modified:
- missions_controller.rb: Refactored mission retrieval logic.
- index.html.erb: Updated mission display logic and added new styles.
- migration file: Added cleanup for invalid missions.
- Various mission JSON files: Updated collections for consistency.
Implements a comprehensive tutorial system that:
- Prompts first-time players with option to take tutorial
- Detects device type (mobile vs keyboard) for appropriate instructions
- Shows interactive steps with objectives for basic controls:
* Keyboard: WASD movement, Shift to run, E to interact
* Mobile: Click/tap to move and interact
- Tracks player actions to progress through tutorial steps
- Saves completion status in localStorage
- Includes polished UI with animations and responsive design
Files added:
- tutorial-manager.js: Core tutorial logic and state management
- tutorial.css: Styled UI components with animations
Files modified:
- game.js: Integrated tutorial check on first load
- player.js: Added tutorial notifications for movement/running
- interactions.js: Added tutorial notification for interactions
- main.js: Imported tutorial manager system
- index.html: Added tutorial CSS stylesheet
Adds a new dialogue option for Kevin where the player can provide a
preliminary security audit update. The feature includes:
- 5 MCQ-style security assessment questions covering:
* Physical security (locks and access control)
* Access control patterns (Derek's suspicious activities)
* Password security (predictable patterns)
* Personnel security (Patricia's firing)
* Data protection (network segmentation)
- Answer tracking system (correct/wrong counts)
- Kevin provides context-appropriate responses to each answer
- Influence point adjustments based on answers
- Available after player makes progress (gets items from Kevin)
- Debrief feedback section that evaluates player's security analysis:
* Excellent performance: 4+ correct answers
* Solid performance: 3 correct answers
* Poor performance: ≤2 correct answers
* Feedback on tradecraft and cover maintenance
This enhances the undercover security consultant role-play and provides
educational value on security assessment best practices.
- Change debrief to be tactical/professional rather than judgmental
- Add emphasis on field immunity for SAFETYNET operatives
- Frame fight as valid tactical choice, not a mistake
- Keep practical implications without moral judgment
- Update mission summary: "Hostile engagement neutralized" instead of "Excessive force allegations"
- Agent 0x99 now validates player agency rather than shaming the choice
- Add objective "defeat_derek_hostile" when Derek becomes hostile
- Emit event "hostile_npc_defeated:derek" when Derek is subdued
- Complete objective when Derek is defeated in fight outcome
- Add "fight" choice alongside arrest/recruit/expose options
- Derek becomes hostile when player chooses to fight (#hostile tag)
- Closing debrief includes moral consequences of fighting vs arresting
- Agent 0x99 discusses excessive force and SAFETYNET protocol
- Mission summary reflects fight outcome with allegations
Fixes two critical issues when multiple doors lead to the same room:
1. Room reloading: Added checks in loadRoom() and createRoom() to prevent
reloading/recreating rooms that are already loaded. This prevents the
room from being unnecessarily recreated when opening a second door to
the same room.
2. Door location detection: Updated removeMatchingDoorSprite() to use
position and direction when finding the matching door. Previously it
only checked connectedRoom, which caused it to remove the wrong door
when multiple doors connected the same two rooms at different locations.
Now it checks:
- Opposite direction matches
- Position matches within tolerance (X for N/S doors, Y for E/W doors)
This ensures that opening different doors to the same room works correctly
without causing room reloads or removing the wrong door sprites.
- Created 9 Ink dialogue files for all mission branches
- Opening briefing with 4-way crisis choice
- Director Morgan, Agent 0x99, and Architect communications
- Branch-specific crisis dialogues for all 4 options
- Closing debrief with deterministic outcomes
- Note: Ink files need compilation fixes before JSON generation
- Added json helper methods for proper escaping
- Use .to_json for multi-line text variables
- Replaced UTF-8 special characters with ASCII equivalents
- Fixed emergency_briefing_text, tomb_gamma_coordinates, mole_evidence_text
JSON now parses correctly but schema validation shows objects need proper format
- Implemented crisis_terminal room with ERB conditionals based on crisis_choice
- 4 distinct crisis scenarios (infrastructure, data, supply_chain, corporate)
- Branch-specific NPCs for each option:
* Infrastructure: Marcus 'Blackout' Chen
* Data: Specter + Rachel Morrow (dual threat)
* Supply Chain: Adrian Cross (recruitable)
* Corporate: Victoria 'V1per' Zhang + Marcus 'Shadow' Chen
- Crisis control systems with unique observations per branch
- Video feed displays showing remote facility status
- Crisis timer tracking progress and countermeasures
- Conditional room type based on crisis choice
- Connection from operations_floor to crisis_terminal
- Set all_flags_submitted when flag4 submitted
Complete scenario structure now includes 7 rooms total
Major design revision to accommodate single floor plan constraint while maintaining
branching narrative and player choice.
Key Changes to README.md:
LOCATION CHANGE:
- Changed from 4 separate physical facilities to single SAFETYNET Emergency Operations
Center (EOC)
- Player remains in SAFETYNET HQ throughout mission
- 4 crisis response zones within same facility
- Player chooses which crisis terminal to directly control
- Other SAFETYNET teams (Alpha/Bravo/Charlie/Delta) visible working in adjacent zones
FLOOR PLAN DESIGN:
- 6 shared rooms (all branches visit these):
1. Emergency Briefing Room (choice presentation)
2. Main Operations Floor (central hub with all 4 crisis zones visible)
3. Server Room (VM challenges)
4. Communications Center (The Architect's taunts)
5. Intelligence Archive (Tomb Gamma discovery)
6. Debrief Room (outcomes revealed)
- 1 branch-specific crisis terminal (player's choice determines which):
7A. Infrastructure Crisis Terminal (power grid remote control)
7B. Election Security Terminal (voter database protection)
7C. Supply Chain Terminal (software distribution security)
7D. Corporate Defense Terminal (corporate zero-day defense)
NPC INTERACTION MODEL:
- ENTROPY antagonists appear via video feeds/screens (not physically present in EOC)
- Maintains distinct antagonists per branch without violating single-location constraint
- Player interacts with remote operatives through crisis terminal displays
FILE STRUCTURE SIMPLIFICATION:
- Changed from 4 separate scenario files to 1 scenario.json.erb with ERB branching logic
- Uses globalVariable "crisis_choice" to conditionally render crisis terminal content
- Dramatically reduces development complexity while preserving 4 distinct narratives
TECHNICAL APPROACH:
- ERB conditionals switch crisis terminal NPCs/objects based on player choice
- Shared rooms identical for all players (easier testing/maintenance)
- VM challenges reusable across all branches
- LORE reveals centralized in Intelligence Archive
- Timer mechanic consistent regardless of choice
BRANCHING PRESERVATION:
- ✅ Still 4 distinct story paths (Infrastructure/Data/Supply Chain/Corporate)
- ✅ Still meaningful choice with permanent consequences
- ✅ Still deterministic outcomes matrix for unchosen operations
- ✅ Still unique antagonists per branch (via video feeds)
- ✅ Still branch-specific technical challenges at crisis terminals
BENEFITS:
- Drastically reduces development time (1 scenario vs 4)
- Easier to maintain consistency across branches
- Simpler testing (test shared rooms once, then 4 crisis terminal variations)
- Cleaner file structure
- More realistic (command center coordination vs field deployment)
Next Steps:
- Update stage 0 planning documents to reflect single-location design
- Begin scenario.json.erb implementation with shared rooms
- Implement choice mechanism in Emergency Briefing Room
- Build crisis terminal templates
Updated Stages 1, 4, 5, and 6 to align with new design:
- Torres knows about casualties and foreign sales (not "journalist" lie)
- Radicalized with extremist ideology over 3 months
- Shows cognitive dissonance but can be de-radicalized
- 5 endings: Turn, Arrest, Combat (lethal/non-lethal), Public Exposure
Changes via small edits:
- Stage 1 (Narrative): Journal entries show radicalization, confrontation dialogue updated
- Stage 1 (Narrative): Replaced "Sympathetic Release" with Combat options
- Stage 4 (Objectives): Updated task 26 to reflect 5 choices, updated stop_upload task
- Stage 5 (Room Layout): Updated evidence correlation matrix and design philosophy
- Stage 6 (LORE): Updated Fragment 1 (radicalization methodology) and Fragment 2 (Architect protocol)
All stages now consistent with evil radicals framing while maintaining moral complexity.
Changes:
- Add Stage 3: Moral Choices with arrest/combat options
- Update Stage 0: Torres knows about casualties, radicalized with extremist ideology (3 months)
- Update Stage 2: Torres is radicalized recruit showing cognitive dissonance, not victim
- Update Mission Summary: 5 endings including arrest and combat (lethal/non-lethal)
Design Philosophy Changes:
- ENTROPY clearly evil radicals (accelerationist extremism)
- Torres radicalized but only 3 months in - can be de-radicalized
- Player has arrest/combat/turn options when confronting ENTROPY agents
- New recruits can be saved before full radicalization
- Maintains player agency across all moral choices
