agentleFS
AshExplainedAshExplained/roblox-skills/roblox-game-ideas/SKILL.md

roblox-game-ideas

Skill5 starsChanged 5 months agoMITView on GitHub

Generate or evaluate Roblox game ideas using retention, social, monetization, mobile-first, and first-session best practices. Use when pitching concepts, choosing a game direction, or improving a core loop.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    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:

    Create Roblox concepts that can win attention and keep players.

    line 10
  2. Distinctive:

    Every strong concept should include:

    line 14
  3. Distinctive:

    Prefer fewer, sharper concepts over many vague ones.

    line 57
  4. Distinctive:

    Ask one question at a time and wait for the answer before the next.

    line 63
  5. Distinctive:

    Stop when the core loop, first session, and monetization are concrete enough to write down without hand-waving.

    line 70
  6. Distinctive:

    Avoid concepts that depend on forced cutscenes, long tutorials, desktop-only controls, hidden progression, harsh early penalties, …

    line 74
  7. Very specific:

    When the concept survives grilling, lock it into a spec with roblox-to-prd, then break it into playtestable slices with roblox-to-issues.

    line 78

≈ “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

Roblox Game Ideas

Goal

Create Roblox concepts that can win attention and keep players. Ideas must be playable quickly, social early, mobile-first, monetizable without breaking fairness, and suitable for frequent updates.

Winner Checklist

Every strong concept should include:

  • First fun action within 10-30 seconds.
  • One clear core loop that repeats cleanly and rewards every cycle.
  • Short-term goals for instant gratification.
  • Long-term goals that can last for weeks.
  • A Day-1 return hook and a visible Day-7 aspiration reward.
  • Social mechanics in the first session, such as co-op objectives, friends, groups, gifting, trading, leaderboards, or shared progress.
  • Mobile-first controls with large targets, forgiving input, and no keyboard dependency.
  • Progression signals: XP, currency, collections, zones, meters, rank, or visible unlocks.
  • Fair monetization woven into the loop through optional passes, boosts, cosmetics, VIP, private servers, or event items.
  • Server-authoritative handling for valuable state such as currency, inventory, purchases, rewards, and progression.
  • A click-worthy icon, thumbnail, and searchable name direction.

Idea Format

For each idea, provide:

  • Title and one-line promise.
  • Player fantasy and target audience.
  • First 30 seconds.
  • Core loop.
  • Social hook.
  • Progression ladder.
  • Retention hooks for Day 1, Day 3, and Day 7.
  • Monetization that avoids pay-to-win.
  • Mobile UX considerations.
  • Update/event pipeline.
  • Main risk and mitigation.

Scoring

Score concepts from 1-5 on:

  • First-session clarity.
  • Core-loop repeatability.
  • Social retention.
  • Progression depth.
  • Mobile fit.
  • Fair monetization.
  • Live-ops potential.
  • Build feasibility.

Prefer fewer, sharper concepts over many vague ones. If a concept scores poorly on first-session clarity or mobile fit, revise it before recommending it.

Grilling Mode (stress-test a chosen concept)

Once the user has picked a direction, switch from generating to grilling — interview them one question at a time to find the holes before any code is written. This is where a fuzzy concept becomes a buildable one.

  • Ask one question at a time and wait for the answer before the next. For each, offer your recommended answer.
  • Walk the design tree in dependency order: core loop → first 30 seconds → progression → economy → social → monetization → mobile controls → live-ops cadence. Resolve each before moving on.
  • Probe with concrete scenarios, not abstractions: "A new mobile player spawns with 0 coins — what is the very first tap, and what does it give them?"
  • Sharpen fuzzy terms into one canonical word the whole project will use ("is it a coin, a gem, or a token? pick one"). Consistent vocabulary now saves confusion in every later skill.
  • Pressure-test against the Winner Checklist and Anti-Patterns above — if an answer implies a long tutorial, desktop-only control, or pay-to-win shortcut, call it out immediately.
  • Capture resolved decisions as you go (in the concept notes or, if the project uses one, CONTEXT.md) so they aren't re-litigated later.

Stop when the core loop, first session, and monetization are concrete enough to write down without hand-waving.

Anti-Patterns

Avoid concepts that depend on forced cutscenes, long tutorials, desktop-only controls, hidden progression, harsh early penalties, pay-to-win shortcuts, or content that exhausts before the 2-3 hour mark.

Next

When the concept survives grilling, lock it into a spec with roblox-to-prd, then break it into playtestable slices with roblox-to-issues. To apply proven mechanics for the chosen genre, use roblox-genre-patterns; to shape the first-session flow, use roblox-game-ux.

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.