agentleFS
EL4CTEOEL4CTEO/roblox-skills/skills/roblox-testing/SKILL.md

roblox-testing

Skill92 starsChanged 47 days agoMITView on GitHub

Testing Roblox code - testable modules with dependency injection, Jest Lua (runCLI, .spec files), legacy TestEZ, running tests in Studio, via Studio MCP, run-in-roblox, or CI with Open Cloud Luau Execution, Lune for pure Luau, multi-client playtests, QA checklist. Use when adding tests, setting up a test runner or CI, or verifying a gameplay change works.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep game rules in pure ModuleScripts (inputs → outputs, no services, no yields): damage formulas, inventory operations, shop pricing, …

    line 10
  2. Very specific:

    Put a jest.config.luau (return { testMatch = { "/*.spec" } }) in each test root.

    line 108

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

Testing Roblox games

Make code testable first

  • Keep game rules in pure ModuleScripts (inputs → outputs, no services, no yields): damage formulas, inventory operations, shop pricing, state machines, serialization, validation. These are trivially testable.
  • Services take their dependencies as parameters (dependency injection) so tests can pass fakes instead of real DataStores, remotes, or Players.
--!strict
-- ReplicatedStorage/Shared/Inventory.luau — pure logic, no Roblox services.
export type Inventory = { [string]: number }

local Inventory = {}

function Inventory.add(inventory: Inventory, itemId: string, amount: number, maxStack: number): (boolean, string?)
	if amount <= 0 or amount ~= math.floor(amount) then
		return false, "invalid amount"
	end
	local current = inventory[itemId] or 0
	if current + amount > maxStack then
		return false, "stack full"
	end
	inventory[itemId] = current + amount
	return true, nil
end

return Inventory
--!strict
-- A service that receives its data store, so tests can inject an in-memory fake.
type Store = {
	GetAsync: (self: Store, key: string) -> any,
	SetAsync: (self: Store, key: string, value: any) -> (),
}

local function newCurrencyService(store: Store)
	local service = {}
	function service.award(userId: number, amount: number): number
		local key = `User_{userId}`
		local balance = (store:GetAsync(key) or 0) + amount
		store:SetAsync(key, balance)
		return balance
	end
	return service
end

-- In tests: a fake store backed by a table.
local fakeData: { [string]: any } = {}
local fakeStore: Store = {
	GetAsync = function(_self, key)
		return fakeData[key]
	end,
	SetAsync = function(_self, key, value)
		fakeData[key] = value
	end,
}
print(newCurrencyService(fakeStore).award(1, 50))

return newCurrencyService

Jest Lua (recommended framework)

Port of Jest (describe/it/expect, mocks, snapshots). TestEZ is legacy; Jest Lua 3 includes a migration guide.

# wally.toml
[dev-dependencies]
Jest = "jsdotlua/jest@3.10.0"
JestGlobals = "jsdotlua/jest-globals@3.10.0"
-- src/shared/Inventory.spec.luau
local ReplicatedStorage = game:GetService("ReplicatedStorage")

local JestGlobals = require(ReplicatedStorage.DevPackages.JestGlobals)
local Inventory = require(ReplicatedStorage.Shared.Inventory)

local describe, it, expect = JestGlobals.describe, JestGlobals.it, JestGlobals.expect

describe("Inventory.add", function()
	it("stacks items up to the limit", function()
		local inventory = {}
		expect(Inventory.add(inventory, "Apple", 3, 5)).toBe(true)
		expect(inventory.Apple).toBe(3)
		local ok, reason = Inventory.add(inventory, "Apple", 3, 5)
		expect(ok).toBe(false)
		expect(reason).toBe("stack full")
	end)

	it("rejects fractional amounts", function()
		expect(Inventory.add({}, "Apple", 0.5, 5)).toBe(false)
	end)
end)
  • Put a jest.config.luau (return { testMatch = { "**/*.spec" } }) in each test root.
  • Runner script (Studio command bar, a test place Script, or cloud task): require(ReplicatedStorage.DevPackages.Jest).runCLI(root, { ci = true }, { root }):awaitStatus() → inspect result.results.numFailedTests.
  • Jest Lua 3 needs the fast flag FFlagEnableLoadModule = true in Studio's ClientAppSettings.json when run locally.
  • Map DevPackages into the DataModel only for test builds (separate Rojo project, e.g. test.project.json).

Where tests run

RunnerUse
Studio (command bar / test Script)Local development.
Studio MCP (execute_luau, start_stop_play, get_console_output)Agents running tests and playtests against the open place — see roblox-studio-mcp.
run-in-robloxCLI that opens Studio, runs a script, pipes output (needs Studio installed; local/Windows/mac CI runners).
Open Cloud Luau ExecutionHeadless CI: runs a script in a real Roblox server for a published place version.
LunePure-Luau modules with no Roblox API dependencies (fast, runs anywhere).

Open Cloud Luau Execution flow (details and a script: references/cloud-test-runner.md):

  1. Build the test place (rojo build test.project.json -o test.rbxl) and upload it as a Saved version of a dedicated test place.
  2. POST /cloud/v2/universes/{u}/places/{p}/versions/{v}/luau-execution-session-tasks with { "script": ... } (API key scope universe.place.luau-execution-session:write; rate limit ≈ 5 tasks/min per key owner).
  3. Poll the task path until state is COMPLETE/FAILED; read output.results and .../logs.

Playtesting in Studio

  • Test tab: Play (you + server), Run (server only, no character), Clients and Servers with 2+ players to test replication, trading, and PvP. Most multiplayer bugs only appear with ≥ 2 clients.
  • Device Simulator toolbar (Test → Device Simulator, top of the viewport): switch phone/tablet/console mid-playtest, with touch-control hints; its Network Simulator applies 3G/4G/Wi-Fi/wired or custom inbound/outbound latency, jitter, and packet loss to test lag compensation and loading/retry UX. Controller Emulator for gamepad input.
  • Streaming: test with StreamingTargetRadius = 64.
  • DataStores in Studio: enable API access and use a separate test universe or data store names.

QA checklist before shipping a feature

  • Works for a player who joins after the feature started (late joiners), and after respawn.
  • Works with 2+ clients; server state is authoritative; UI updates on all clients.
  • Player leaving mid-action (trade, purchase, round) leaves no broken state or leaked tables.
  • Remotes rejected when spammed or sent malformed arguments.
  • Data saves and reloads (rejoin), including teleports between places.
  • Mobile, gamepad, and keyboard all work; UI readable on a phone.
  • No errors/warnings in Output (server and client).

Related skills

roblox-tooling (CI, Rojo), roblox-studio-mcp (agent-driven playtests), roblox-open-cloud, roblox-code-review.

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.