Vibe Tycoon: The Story from v0.99.0 to v1.3.1
Vibe Tycoon’s move from v0.99.0 to v1.3.1 was the point where the project stopped feeling like a collection of connected features and started behaving like a real game product.
The previous releases had already established the core Prompt loop, the studio economy, agents, upgrades, Growth, Prestige, rewards, and mobile navigation. The next challenge was different. The game needed to become easier to understand, more satisfying to play, more stable under real interaction, and more disciplined internally.
This period was about turning the foundation into a product with a stronger opening, clearer feedback, a consistent visual language, better economy control, cleaner architecture, and more honest failure handling.
The work was not a straight line. Some changes improved the player experience immediately. Others exposed deeper problems: rewards that were technically implemented but hard to find, an economy that could run away under aggressive hiring, a large GameScreen file that had become difficult to reason about, crashes caused by update loops, controls that looked finished but were not fully accessible, and a release process that needed stronger evidence.
The result was a sequence of releases that progressively tightened the game from every direction.
v0.99.0: Starting from a Stronger Foundation
Version v0.99.0 was the starting point for this chapter.
The release reworked progression, Growth, Prestige, milestones, agent guidance, and mobile navigation. AI Operations received clearer plans, purchase guidance, upgrade-readiness cues, and Product Studio locked-state guidance. The selected AI Operations plan was persisted, while prompt jobs, Cash chips, compact mobile behaviour, bottom navigation, click feedback, and congratulations timing were refined.
This release established the problem to solve next: the game had the systems, but the systems needed to become more discoverable and more satisfying. Players needed to know what to do, what they had earned, what they could unlock, and why the next purchase mattered.
v1.0.0: Making Rewards Findable
Version v1.0.0 focused on reward visibility and usefulness.
The Hackathon sprint boost was changed so it increased both Cash and AI-token throughput by 2x for five minutes, instead of affecting Cash alone. More importantly, the boost was moved into a visible Home action bar that stated the exact reward before the player tapped it. A valuable system was no longer buried inside Upgrades > Tool Bench.
The token rescue was renamed Sponsored Top Up. The new wording made it clear that it was a sponsor offer, what reward it provided, and where the ad was involved. The game began explaining its economy directly in the top bar as well:
- Cash comes from work and product payouts.
- AI Tokens come from AI Operations and Sponsored Top Up.
- Reputation comes from completed jobs and milestones.
The low-Cash Bridge Round was also made more useful as the company grew. Rather than stopping at a fixed reward ceiling, it scaled with the gap to the player’s next purchase, capped at twice the current Cash balance.
The breakthrough in v1.0.0 was simple but important: rewards are only useful if players can find them and understand them. The release moved Vibe Tycoon away from hidden mechanics and toward visible, explained systems.
v1.1.0: Cleaning Up the Product Behind the Product
Version v1.1.0 combined player-facing improvements with a major internal cleanup.
The Trophy Shelf became a full-screen overlay that could be opened from the Home trophy icon, Milestones, or an unlock notice. Players could tap a trophy and reach its related product card directly.
Congratulations cards became immediately responsive by removing the timed interaction lock. The interface no longer made the player wait after an event had already finished.
The bigger change was architectural. GameScreen.tsx had grown to approximately 4,814 lines. It was split into a 1,292-line core plus eleven feature files. Source-text tests were converted into behaviour tests, making the test suite more concerned with what the game did than with the exact wording or structure of its implementation.
Approximately 4,200 lines of dead code, orphaned assets, and one-off scripts were removed. Project instructions were reduced and moved into proper documentation. Jest mocks were consolidated, an unused echarts dependency was removed, and new development commands were added for fresh web builds, web checks, and the broader gate.
The headless playtest gate was brought back to green against the current UI. A Settings gear hitbox defect was identified and recorded rather than quietly ignored.
This was not just tidying. It was a response to a real development problem: when the interface, tests, scripts, and source structure all describe slightly different versions of the game, every new feature becomes harder to verify. v1.1.0 reduced that drift.
v1.1.1: Stopping the Economy Runaway
Version v1.1.1 was a focused economy balance release.
The problem was that hiring became dangerously cheap from the tenth agent onward. A player could rush-buy a large roster, reach billions of Cash, and achieve a full Prestige in minutes. The early game was working correctly, so the fix had to preserve the first-hire and first-job experience while preventing late-game runaway growth.
Hiring costs were made sharply more expensive from the tenth agent onward. The early curve remained untouched, protecting the opening experience while restoring control over the long-term economy.
An internal idle-game mathematics gap analysis was also added under docs/research/. The release deliberately avoided unnecessary save-schema, progression, or monetisation changes.
The lesson was that balance problems are often invisible during the first few minutes of play. A system can feel generous and exciting early on while still destroying the entire progression curve later. v1.1.1 kept the beginning fun without allowing the back half of the game to collapse.
v1.2.0: The Game Started Feeling Like a Game
Version v1.2.0 was the major visual and interaction pass.
The top bar was rebuilt around a stage card with a pixel progress bar and Prestige Points. Resources moved into three icon-first pills, and Settings became a clear square button rather than an ambiguous control.
A full-width news ticker was added to carry industry and studio headlines, milestone updates, and incident alerts in one consistent band. New companies now ask players to name their studio, making the company feel owned rather than generic.
PAPI began announcing unlocks, new tools, and level-ups through one pixel dialog at a time. These announcements no longer replayed after reload, which removed a major source of friction and confusion.
Agent interaction also became more meaningful. Players could tap an agent, open a conversation, and choose a focus. Every focus selection received an answer, including cases where two agents already covered the same area. Sales Skill and Local AI were removed from the focus choices where they had no actual gameplay effect.
Milestones were reorganised so claimable rewards appeared first, followed by the closest active goals with progress bars. Claimed goals folded away instead of continuing to occupy attention.
The Tool Bench was rebuilt as compact rows with an x1 / x10 / All purchase header. AI Operations received a clearer agent roster, a hire button above the roster, per-agent multiplier choices, and a single multiplier-cycle control.
Settings, Help, Product Studio launch cards, and bottom navigation were cleaned up. The active tab became visible, and pages stopped disappearing underneath the dock.
This release also dealt with real failures. The Settings gear hit target was fixed. Decor editing and Android Back behaviour were corrected. Token rate calculations were aligned with the Prompt loop. Roster faces were made to match the correct desks. Two Maximum update depth exceeded crashes were fixed, one in the starfield and one in the Growth sampler.
The important breakthrough in v1.2.0 was that the game began to communicate in a single visual language. The pixel treatment was no longer limited to isolated assets. It became the shared kit for the shell, cards, tabs, switches, sliders, dialogs, resource displays, and navigation.
v1.3.0: A Full Game-Feel Pass
Version v1.3.0 was described as the game-feel pass, and that is exactly what it became.
The title screen gained a centred logo fade, loading feedback inside the Start button, and a transition from Start into the player’s studio. The first job received a bouncing TAP! prompt and the instruction changed to:
Tap the screen to prompt
That single change made the opening action much clearer. Instead of asking a new player to understand the whole system, the game now gave them one obvious first move.
Home received a full-width navy bottom bar with labelled tabs. The open tab visibly popped up, while Daily, Decor, Trophies, and PAPI became raised action buttons. The terminal gained a Hide/Show control, and the news ticker became the top strip of the home surface.
Pages were given more room. Upgrades and AI could slide the stage card and news ticker away when necessary. Upgrades, AI, Product Studio, Growth, and Settings moved onto raised game cards with pixel tabs, game-style switches, and sliders.
AI Operations gained a Token flow meter showing supply against burn. Product Studio displayed the three portfolio slots and a locked teaser supported by a Reputation bar. Decor and Rooms became bottom sheets with Shop and Owned tabs. The next room showed its Reputation progress, and Reset required confirmation before wiping the layout.
Help buttons became real explanatory dialogs that could not be covered by other surfaces. The AI Operations plan cards were corrected so the displayed price matched what the first tap actually purchased. Cash-rate calculations were corrected to include product income and company-timer payouts. Roster Choose now opened the agent chat on the focus grid, and an empty Trophy Shelf directed players toward Product Studio.
The v1.3.0 breakthrough was onboarding and feedback. The game stopped assuming that the player already understood it. It demonstrated the first action, exposed the relationship between resources and systems, and gave every major screen a shared visual grammar.
v1.3.1: From Game Feel to Reliability
Version v1.3.1 moved the focus from presentation into hardening.
Upgrade price buttons were turned into Cash progress bars, giving players an immediate view of how close they were to the next purchase.
The release also surfaced a much larger reliability agenda. Persistent mutations across Cash, Tokens, Reputation, XP, agents, jobs, tools, and automation needed to be audited for crash safety. Reward settlement needed an exactly-once path so job payouts, Reputation, Tokens, XP, product revenue, task claims, wheel rewards, Prestige Points, and offline earnings could not be duplicated or lost.
Agent and job lifecycle recovery became an explicit concern across the complete chain: offer, accept, assign, run, incident, recover, review, ship, and release. Stuck jobs needed detection, timeouts, deterministic recovery, and age-based handling.
Runtime failure recovery and loading watchdogs were added to the hardening direction for saves, loads, settlements, purchases, upgrades, job completion, product launches, Prestige, automation, and offline returns.
The release also addressed player-facing explanations. Disabled controls and dead buttons needed to explain why an action was unavailable, what value was required, and what was missing. Other cleanup work included removing retired labels such as “JOB BOARD” and “NEXT UP”, replacing raw Founder Tree IDs with player-facing names, fixing short-phone overlap in the Home action rail, and improving the PAPI store explanation of faster jobs.
Further UI work targeted the phase header, bottom navigation, Home action rail, Tool Bench, AI Operations, Settings, Help, Product Studio, currency CTAs, the terminal surface, and the Welcome Back panel.
The release also continued the crash investigation from v1.2.0. The starfield’s 42-node animation loop had been flooding React with a state update every frame, while the Growth-history sampler was calling setState every second and writing to storage inside its state updater. Both patterns could produce Maximum update depth exceeded failures in React Native and Fabric.
v1.3.1 represented a change in maturity. The project was no longer satisfied with a feature existing. It needed to survive reloads, crashes, repeated settlement, incomplete jobs, Android lifecycle changes, and player actions taken at exactly the wrong moment.
The Challenges Behind the Releases
The changelogs show the visible outcomes, but the real story is the problems underneath them.
The game kept outgrowing its old descriptions
The player-facing language repeatedly fell behind the mechanics. Old labels remained after the underlying system had changed. Screens described retired concepts. Buttons suggested a value that was not actually being delivered. The solution was not simply copy editing. The interface had to be audited against the mechanics and rewritten around what the game really did.
Features could be built but still not be reachable
The Sponsored Top Up and other reward surfaces demonstrated an important release lesson: code can exist without a player ever reaching it. Mounting, navigation, surface conditions, and real interaction all had to be checked. This is why later releases placed more emphasis on headless playtests, runtime gates, and evidence rather than source inspection alone.
The economy needed to be fun without becoming meaningless
The economy had to reward activity, preserve momentum, and make upgrades feel valuable. But if hiring costs or reward curves were too generous, the player could skip the intended progression and reach Prestige almost immediately. v1.1.1 showed the value of protecting the early experience while adding late-game controls.
Web behaviour and Android behaviour were not identical
A screen that looked acceptable on web could still break on Android. Currency cards, Back handling, native rendering, modal placement, background assets, bundle size, and crash behaviour all had to be tested against the actual mobile environment.
The codebase needed to keep pace with the game
The GameScreen split and removal of thousands of lines of dead code were not cosmetic refactors. They reduced the cost of understanding and testing the game. Without that work, every new feature risked adding another layer to a surface that was already difficult to reason about.
The Breakthroughs
Across v0.99.0 to v1.3.1, the main breakthroughs were:
- The Prompt loop became visible. Players could see what to do, what they earned, and how agents contributed.
- Rewards became findable. Important boosts moved into visible action surfaces with explicit outcomes.
- The economy became more honest. Cash, Tokens, Reputation, product income, offline earnings, and Prestige were explained and connected to real rules.
- The UI gained one visual language. Pixel cards, tabs, switches, sliders, dialogs, bars, and navigation began to feel like one game rather than separate screens.
- The opening became understandable. A new player receives a clear first action instead of being dropped into a complex studio dashboard.
- The architecture became maintainable. GameScreen was split, dead code was removed, mocks were consolidated, and behaviour tests became more meaningful.
- The game became safer to operate. Save integrity, exactly-once rewards, job recovery, loading watchdogs, and runtime failure states became explicit engineering concerns.
The Wins
The wins were not only new screens or new features. They were moments where the product became more truthful:
- A player could see why an upgrade was affordable or not.
- A reward explained exactly what it would provide before the tap.
- A new player received a clear first action.
- A studio could be named instead of remaining anonymous.
- An agent focus choice produced an answer instead of silently doing nothing.
- A claimed milestone stopped taking up attention.
- A page stopped hiding underneath the bottom dock.
- A mobile control became large enough and clear enough to use.
- A crash caused by runaway state updates was identified at its actual source.
- A large screen was split into understandable feature boundaries.
- An economy exploit was closed without damaging the first-hire experience.
- A release gate tested the current game instead of an obsolete version.
The Whole Story
The movement from v0.99.0 to v1.3.1 was the story of Vibe Tycoon becoming more deliberate.
At v0.99.0, the game had a strong foundation but still needed to tell players what mattered. v1.0.0 made rewards visible. v1.1.0 made the codebase and testing process healthier. v1.1.1 protected the economy from runaway progression. v1.2.0 gave the game a consistent visual language and more meaningful interaction. v1.3.0 made the first minute and the major pages feel like one product. v1.3.1 began hardening the whole system so it could survive real use rather than only happy-path demos.
The project’s biggest breakthrough was understanding that game quality is not just the sum of mechanics. It is the agreement between the mechanics, the words, the visuals, the economy, the save system, the mobile runtime, the tests, and the release process.
Vibe Tycoon became stronger every time those layers were brought back into alignment.
The story is not that every problem disappeared. The story is that the project became better at finding problems, explaining them, fixing them at the root, and proving that the fix reached the player.
That is what the journey from v0.99.0 to v1.3.1 represents: the transformation of a working idle-game foundation into a more readable, more intentional, more stable, and more complete studio-management game.
Log in or sign up for Devpost to join the conversation.