Lifting Spirits is a cozy, vertically scrolling platformer with six levels. Climb a tree, meet its spirits, and collect items to reach the top and reunite with your friend.
Our game ranked #3 overall out of 209 games for the 2026 Spring Comfy Jam. We also ranked #2 for Polish, #7 for Art, #7 for Limitation, and #10 for Audio.
Genre: Platformer
Platform: Browser
Team Size: 5
Timeline: 2 weeks
Engine & Tools: Godot, Dialogic 2, Figma, Excel
My Contributions: Barks, Dialogue, Flavor Text, Worldbuilding, Lore, Shaping Character Design, Storyboarding, Narrative Direction
About the Game and Restrictions
Our game had to feature a seamless loop and incorporate at least one theme: strawberries, bees, or flowers. We built the loop around repeatedly climbing a tree, revealing new platforms and layouts with each ascent to increase the difficulty.
I wanted the repetition to have an emotional purpose: the power of persistence. By continuing to climb and show up for her best friend Clover, the player helps her break out of a spiral of feeling like she isn’t good enough. The disgruntled spirits we meet are also inspired to change with each climb.
I Started With W
orldbuilding
If the player was going to climb a big tree, it needed to feel ancient, meaningful, and grand.
Because the game was going to have 6 levels, each section of the tree should also feel different and contribute to the overall fantasy world.
I wanted the tree to feel somewhat realistic, with spirits in each section reflecting the organisms that would naturally live at that height.
The game artist and I brainstormed the world and its inhabitants during a one-hour workshop in FigJam.
I came to the meeting with an environment one-pager for our Great Tree in Figjam, which we edited live together as we nailed down the characters we’d create and the ‘micro-world’ of each level.
We left the meeting with clearer character concepts and a shared vision for how each level should look and feel distinct.
I also incorporated all lore and history sections into our Game Design Document for everyone’s reference.
Worldbuilding in Level Environments
Level 1 is at the trunk of the tree, so mushrooms and roots made the most sense with darker colors.
Level 2 is where flowers and branches shine, so a flower spirit would fit here and our jam’s theme, too.
Level 3 is near the top of the tree and should look more somber to show how The Great Tree was decaying and getting sicker. We also implied a little bit of a ‘class divide’ by having more humanoid characters like gnomes and fairies live at the top in houses, while more plant-like characters lived lower on the tree. (Implied classism!?)
I Also Created Lore to Help Shape Character Designs
I developed lore around the game’s ‘Spirit Guardian,’ who was the central conflict that motivates players to keep climbing the tree and keep playing our game’s loop.
I initially explored a friendly rivalry theme between the protagonist and her best friend where they could each represent different archetypes and outlooks, like an introvert who goes with the flow vs. an extrovert who people pleases and easily burns out.
But I noticed Clover’s character design was starting to look too ‘evil’ for a cozy game jam. I didn’t want our people-pleasing rival to be seen as a villain, nor to make the story more complex than it needed to be. This was a platformer, so the story’s conflict should be relaxing, not high stakes.
So we decided as a team to drop the rivalry theme and instead keep them simply as best friends and softened Clover’s look and feel.
Early Character Design Concepts
Final Character Concepts
Then I Started to Write Linear Dialogue
Because we had only two weeks for the jam, I chose Google spreadsheets to rapidly prototype bite-sized dialogue and make its flow easy for my teammates to follow the story’s flow and plot progression, all through dialogue lines. I used arrows (‘→’) to mimic the ‘Continue’ button and show how dialogue would be chunked.
I also used bold text to cue players in on key information, similar to the Ori games. This would also remind me which words to bold directly in the game engine.
Instead of writing longer character bios, I used the Spirit Type and Mood and Tone sections so that our Artist could have some guidance when concepting the characters. Our Sound Designer also used this to develop specific sound effects and ‘theme songs’ for the levels they appeared in.
The dialogue that was implemented into the game is below:
Implementing Feedback From Teammates and Playtesting
My dialogue had gone through 3 revisions: one initial pass between my Game Designer and Artist; another with friends and family playtesting; and then one final pass with the Game Designer and Artist again before submitting our game to the jam.
Coming from a UX Writing background where every piece of UX content on the interface must be crystal clear, I felt like my first draft of the dialogue was too explanatory of the world’s customs and how those characters were feeling.
I also felt like my dialogue held a lot of restraint and I didn’t use a lot of dialogue boxes—because in tech and American consumer apps, space is precious. In other words, I was used to the principle that “less is more.”
I was also heavily relying on puns for humor (ex. staying put as a rock spirit, twig snapping under pressure, etc.)
But showing restraint actually ended up working against me!
My game designer, artist, and play testers ended up with more questions than answers. They wanted to understand why the player character couldn’t speak and what the player’s objective was in the opening scene.
The opening dialogue had to strike a balance between avoiding over-explanation and clearly establishing the player’s purpose in this world—and why they mattered.
This feedback helped me show, rather than tell, why our player character was mute in the second level (shown in the GIF below).
It also helped me tighten the first character’s dialogue to establish the state of their world and what the player character’s role is in it.
Implementing Dialogue in Godot
During the playable prototype phase, I was already implementing WIP dialogue into the game engine.
I used BBL formatting to bold key words and phrases from the spreadsheet, and also used text-speed modifier code to control how quickly or slowly a character spoke and delivered their lines.
This could emulate voice and tone when we didn’t have time to implement different SFX for each character’s voice.
For example, I used this technique when Fauna tries to insult the player character and make them feel insecure.
There’s a brief pause, represented by an ellipsis (‘…’), that’s paced out to imply that Fauna is waiting for the player to react to her mean comment.
And when a response doesn’t come, Fauna gets flustered in true tsundere fashion.
This helps immerse the player in the characters and in the world.
Writing Flavor Text for Items
I created a tab in the dialogue spreadsheet that included flavor text for each item the player collects.
I defined which column was internal usage and which were okay for in-game usage so that our Game Designer and Programmer would know which item did what.
Although we didn’t have time to implement the flavor text, it was designed to appear when players hovered over the items in the game.
The only item that had in-game benefits was the Berry Boon, which increased the player’s stamina bar so that they’d be able to get to the top of the tree, which they couldn’t do in the first level (it was impossible).
In a longer game, the other items could be used for other power-ups!
Creating a Storyboard to Visualize Story Beats and
Narrative Flow at Each Level
I created a FigJam flowchart for our game designer/level designer, mapping how story beats connected to each level and how the game’s falling animation tied everything together.
Characters appeared at the start of each level to give players a break between sections of active gameplay, balancing action-oriented play with quieter, more passive moments.
My diagram became a key storyboard that the game designer, game artist, and programmer all used to help implement the pacing and sequencing of the game!
Directing the Final Cutscene
An idea I’m really proud of pitching and convincing my team to implement was on the final cutscene of the game.
In our second iteration, Clover said, ‘Would you like to join me for tea?’ before the scene faded to black, with the player character standing still, far away from her.
I argued that if the player character and Clover were truly best friends—and we wanted an emotional payoff to our theme of persistence—they should walk toward the tea table together before the scene slowly faded to black.
I think our final cutscene really tied all the loose ends together and made players feel satisfied that the two friends were finally reconnected!
What I Would Do Differently Next Time
Improve the climbing UX instructions at the beginning of the game
While we taught players that the spacebar advanced dialogue, we should’ve added a line to the next tooltip explaining how to climb: “Hold Space AND a Key.” When I showcased the game at a local game dev meetup, I noticed players would hold the spacebar without realizing they also needed to press the Up Arrow or S key. Although they eventually figured it out, clearer instructions upfront would’ve made the mechanic more intuitive.Remove thorns from bush assets to preserve the cozy tone
Later levels include thorny bushes that block the player’s path. We originally planned for the thorns to “prick” the player and immediately make them fall, but ultimately decided that mechanic felt too violent for the cozy tone of the jam. I still would’ve preferred removing the thorns from the bush assets entirely to better maintain the game’s calm and cozy atmosphere.Change the environment backgrounds with each ascent to show The Great Tree deteriorating
Due to time constraints, we kept the main environment backgrounds consistent across climbs and instead varied platform assets to represent different biomes and different gameplay mechanics. However, since the story centers on The Great Tree getting sicker, changing the backgrounds could’ve visually communicated that progression—for example, by having the tree trunk gradually fade, grey, or rot. This would’ve reinforced the story’s urgency and would’ve made the player’s goal of healing the tree feel more impactful.