Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add furkantokkan/agent-foundry --skill game-feel-polishgit clone --depth 1 https://github.com/furkantokkan/agent-foundryWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/furkantokkan/agent-foundry/game-feel-polish)<a href="https://agentmods.dev/skills/furkantokkan/agent-foundry/game-feel-polish"><img src="https://agentmods.dev/badge/skills/furkantokkan/agent-foundry/game-feel-polish/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/furkantokkan/agent-foundry/game-feel-polish"><img src="https://agentmods.dev/badge/skills/furkantokkan/agent-foundry/game-feel-polish.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00124 | $0.02294 |
| Opus 5.5 | $0.00050 | $0.00918 |
| Sonnet 5.5 | $0.00025 | $0.00459 |
| Haiku 4.5 | $0.00012 | $0.00229 |
Grade A, and why
game-feel-polish scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 26d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Game Feel Polish
Polish is feedback layered on top of a committed gameplay result. This skill
gives an agent one architecture for that layer and the numbers that make it
read as impact instead of noise. The systems are adapted from Miisan's Liquid
UI Kit for Godot 4 (MIT, see this skill's THIRD_PARTY_NOTICES.md); the
Unity workflow uses UI Toolkit for new screen-space widgets. This is an
implementation skill and reference kit, not a precompiled Unity package.
Use team-polish when the request is a multi-role release-hardening pass.
When More Mountains Feel is installed, use its pinned project documentation
and existing players; do not require an unbundled companion skill.
Rules that hold in every engine
- One hub, one switch. Every effect is a call on one hub (
Juiceautoload in Godot, one service in Unity). The hub ownsenabledand a 0 to 2intensitymultiplier. Turningenabledoff must reset every owned effect: release time requests, camera offsets, screen pulses, transient labels, voices and rumble. Preserve pause and other owners' state. Keep essential notifications visible. Widgets work when the hub is missing; they only skip sound and shake. - Feel never decides gameplay. Fire feedback after the committed result. A refused action (a disabled button) gets a refusal shake and an error sound; state does not change.
- Trauma, not "shake for N seconds". Overlapping hits add trauma
(clamped 0 to 1); the camera uses
trauma ^ 2so small hits stay subtle and big ones slam. Decay by real time, not by scaled delta, so the shake keeps moving during hitstop. - Wall-clock time for time effects. Hitstop and slow motion measure their own end with the system clock; a freeze timed with scaled delta freezes its own countdown. Hitstop beats slow motion. A longer hitstop extends a running one; a shorter one never cuts it.
- Remember the rest state. Tween helpers record the node's resting scale, rotation, position, and modulate the first time they touch it, kill the previous tween on that channel, and always return to rest. Spamming a punch can never leave anything stuck at the wrong size.
- Fast out, slow settle. Impact tweens spend about a fifth of their time going out (quad ease-out) and the rest coming back (elastic ease-out). An even in and out reads as a wobble, not an impact.
- Separate presentation channels. Use polygons, text and palette colours where custom drawing helps; give hover, press, intro and refusal separate state. Unity USS supports transitions: retain native styling, text and input semantics rather than replacing everything with custom drawing.
- Sound is randomised and de-duplicated. Each sound is a bank of several takes; never repeat the previous take, vary pitch about 0.92 to 1.08, and drop a second trigger of the same bank inside 25 ms. Two identical samples on one frame double the amplitude and clip.
- Pool what fights spawn. Damage numbers, audio players, and toasts are pooled or capped. Allocation on the hit frame is the stutter players notice.
- Scale everything to impact strength. One
impact(position, strength)call maps strength to shake, hitstop, flash, aberration, zoom punch, rumble, and the sound bank with tuned ratios. Screen-wide effects open only above a threshold so routine hits do not wash out big ones. - Prove it with A/B. Ship a toggle that disables all juice at once and compare the same screen with and without. If an effect makes no visible difference in that comparison, delete it.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 26d ago Changed 790a6cdb0e5a
- 1mo ago First seen · 183 lines · 124 tokens per session scan A 254fb183c8c9
game-feel-polish is a skill published in the GitHub repository furkantokkan/agent-foundry (11 stars, last pushed 2d ago), licensed MIT. It adds 124 tokens to every session and 2,294 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-25.
Other skills, from other repositories
lens-review
Review a design artifact (GDD / vision / ADR / mechanic spec / prototype doc / any design document) through 100 game design lenses. Auto-picks 3-5 most relevant lenses based on artifact type, outputs multi-angle insights + multiple directions with explicit trade-offs (no single prescribed answer). Hedged tone, not…
ux-design
Guided, section-by-section UX spec authoring for a screen, flow, or HUD. Reads game concept, player journey, and relevant GDDs to provide context-aware design guidance. Produces ux-spec.md (per screen/flow) or hud-design.md using the studio templates.
ux-review
Validates a UX spec, HUD design, or interaction pattern library for completeness, accessibility compliance, GDD alignment, and implementation readiness. Produces APPROVED / NEEDS REVISION / MAJOR REVISION NEEDED verdict with specific gaps.
art-bible
Guided, section-by-section Art Bible authoring. Creates the visual identity specification that gates all asset production. Run after /brainstorm is approved and before /map-systems or any GDD authoring begins.
unity-ui-patterns
Unity UI/UX pattern design-to-code translation. Screen flow architecture, View/ViewModel separation, HUD architecture, feedback & juice systems, dynamic list/grid views, transition & animation contracts. UI Toolkit only. DESIGN INTENT format: INTENT/WRONG/RIGHT/SCAFFOLD/DESIGN HOOK. Based on Unity 6.3 LTS.
unity-ui
Unity 6 UI development guide. Use when building user interfaces, menus, HUD, buttons, or any UI elements. Covers UI Toolkit (recommended for new projects — USS, UXML, UI Builder, data binding), uGUI/Canvas (legacy runtime UI), and IMGUI. Based on Unity 6.3 LTS documentation.