agentleFS
rodrigo-guerraarodrigo-guerraa/roblox-hub/skills/vendored/game-feel/SKILL.md

game-feel

Skill2 starsChanged 31 days agoMITView on GitHub

Engine-agnostic game-feel ("juice") design guidance, adapted for Roblox. Covers layered feedback, decaying-trauma camera shake, faked hit-stop, eased tweens, squash and stretch, importance tiers and accessibility toggles. Use when a Roblox action (dig, hatch, collect, buy, rebirth, hit, jump, pickup) works but feels flat or weightless, when adding TweenService pops, ParticleEmitter bursts, sounds, camera shake, flashes or coins flying to a counter, or when deciding how much feedback an event deserves. It adds polish on top of a working mechanic; it does not implement the mechanic.

9 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use camera-systems, procedural-gen and 7 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • ukanwatSkill

    camera-systems

    Build game cameras that feel good — 2D follow with a deadzone, look-ahead, smoothing, and level-bounds clamping; 3D third-person orbit with collision and first-person look; plus multi-target framing and a shake hook. Engine-neutral techniques that pair with the engine's camera node and rigs like Unity Cinemachine or Godot Camera2D/PhantomCamera. Use when the user mentions camera follow, follow camera, deadzone, look-ahead, camera smoothing, camera bounds/ limits, third-person camera, orbit camera, first-person look, Cinemachine, or camera jitter.

    Also in 4 projects that use this

  • ukanwatSkill

    procedural-gen

    Generate game content procedurally — seeded deterministic RNG, value/Perlin/ Simplex noise for terrain and heightmaps, grid dungeon generation (rooms + corridors, BSP, random walk), and weighted loot/drop tables. Engine-neutral algorithms. Use when the user mentions procedural generation, perlin/simplex noise, random seed, dungeon generator, heightmap/terrain, or loot tables.

    Also in 4 projects that use this

  • MRCalderon3DSkill

    unreal-enhanced-input

    Implement Unreal input with Enhanced Input — Input Actions, Mapping Contexts, triggers/modifiers, and runtime remapping.

    Also in 4 projects that use this

  • +6 more from ukanwat

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep juice off the simulation.

    line 33
  2. Distinctive:

    Verify by triggering each event repeatedly.

    line 35

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Game feel (juice)

Where this conflicts with the roblox-lessons skill, roblox-lessons wins.

The difference between a mechanic that works and one that feels good is feedback: a layered, briefly exaggerated response to each action.

On Roblox

  • Client-only, on server-confirmed events. Play the payoff when the server's push arrives (a RemoteEvent, an attribute change, a coin delta), not on the button press. Give the press a neutral response, so a refused action never celebrates.
  • Tweens: TweenService:Create(obj, TweenInfo.new(t, Enum.EasingStyle.Back, Enum.EasingDirection.Out), goal). GUI pops go through their own UIScale, never one shared with fit-to-screen scaling. Models tween a value that drives PivotTo.
  • Particles: ParticleEmitter:Emit(n) from a pooled, disabled emitter. Small additive particles vanish in daylight and inside avatars (roblox-lessons assets-visuals.md).
  • Sound: route through SoundGroups and vary PlaybackSpeed slightly. See audio-design.
  • Camera shake: offset Camera.CFrame on the client after the default camera runs (BindToRenderStep at Enum.RenderPriority.Camera.Value + 1), or use Humanoid.CameraOffset. Never move the character.
  • Hit-stop: Roblox has no global time scale, and you can't pause physics for one player. Fake it locally: AnimationTrack:AdjustSpeed(0) for 50 to 120 ms, then restore.
  • Flash: a Highlight, or a SurfaceAppearance.Color tint on textured meshes, tweened back.
  • Screen effects (coins flying to the counter, floating text, full-screen flash) go on a high-DisplayOrder ScreenGui with IgnoreGuiInset.

Core principle

One satisfying event is 5 to 8 tiny responses within about 100 ms: a sound, a burst, a flash, a small shake, a number pop and a counter bump. Exaggerate briefly, then return to rest. Scale the response to the event's importance: a coin tick is not a rebirth.

Workflow

  1. Confirm a client-visible event hook exists. If it doesn't, add it first.
  2. Pick 2 or 3 feedback channels per event. Add more until the event reads clearly, then stop.
  3. Ease everything: Back Out for a pop, Quad or Cubic Out for settling, Sine InOut for camera moves. Linear motion feels robotic.
  4. Save shake and hit-stop for impacts. Keep them short and scaled, and never use them on routine actions.
  5. Keep juice off the simulation. Shake the camera, not the body. Hit-stop is purely visual.
  6. Define small, medium and large presets, and assign every event to one.
  7. Verify by triggering each event repeatedly. Check that it fires, decays to rest, never blocks input, and reads on a phone.

Pitfalls

  • Moving the HumanoidRootPart for shake breaks physics and replication.
  • A fresh math.random offset every frame buzzes. Multiply a decaying trauma value by math.noise or summed sines instead.
  • Permanent exaggeration (a scale that never resets, a shake that never decays) stops reading as feedback.
  • Over-juicing routine actions hides real impacts and causes nausea. Offer reduce-shake and reduce-flash options.
  • Long freezes or uncancellable sequences block input.
  • Effects spawned per event without cleanup leak. Use pools or Debris:AddItem.
  • Celebrating before the server confirms, then the action fails.

References

  • references/feedback-recipes.md: the trauma math with a Luau sketch, an easing cheat sheet using Roblox EasingStyle names, the rest of the feedback menu, importance tiers, accessibility, and Roblox bindings.

Related: audio-design, game-ui-ux, roblox-lessons.


Adapted from awesome-gamedev-agent-skills (Apache-2.0). Modified; see SOURCE.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.