ashexplained/roblox-skills

roblox-to-issues

Break a Roblox game PRD, roadmap, or plan into small independently-playtestable vertical slices instead of building the whole game at once.

查看源码

使用方法

  1. 复制安装命令,在自己的智能体环境安装前先检查源码包。
  2. 从原始文档核对所需工具、凭据和支持的宿主。
  3. 说明输入材料与期望交付物,并对照来源检查输出。
npx skills add https://github.com/ashexplained/roblox-skills --skill roblox-to-issues

开始前需了解

  • 安装不包含模型订阅、API 额度或访问文件的权限。
  • 仓库星数和安装量反映活跃情况,不等于独立质量或安全测试。
仓库原始内容

按源仓库内容呈现,保留标题、案例、代码、表格、链接以及原文引用的演示图片。

Roblox → Issues (Vertical Slices)

Break a plan into small issues that are each buildable and playtestable on their own. This is the antidote to "build the whole game in one go, everything half-broken."

The tracker location and label vocabulary should have been provided to you in CLAUDE.md / docs/agents/. If not, run setup-roblox-skills. Slices publish either as GitHub issues (gh) or as markdown files under roadmap/<feature-slug>/slices/ — whichever the setup configured. In all cases, keep ROADMAP.md as the human-readable index.

What a Roblox vertical slice is

A tracer bullet: the thinnest end-to-end path that a player can actually feel, cutting through every layer it touches — client input, server authority, remotes, DataStore, and UI — not one horizontal layer.

  • Right (vertical): "Player taps the ore, server grants 1 coin, HUD shows the new balance, balance survives rejoin." Touches input + remote + server + data + UI. Playable.
  • Wrong (horizontal): "Build the entire economy module." / "Build all the UI." Nothing is playable until everything is done — exactly the failure we are avoiding.

Slice rules

  • Each slice delivers one narrow but complete player-facing capability.
  • Each slice is verifiable in a Studio playtest on its own (this is the acceptance signal).
  • Prefer many thin slices over a few thick ones. If a slice can't be playtested in isolation, split it.
  • Valuable state (currency, inventory, purchases, rewards, progression) is server-authoritative from the first slice — never a client-trust shortcut you plan to fix later.
  • The first slice is the tracer bullet: the smallest thing that proves the core loop end-to-end (one action → one reward → one visible next goal). Build it before anything else.

Slice types

  • AFK — can be built and verified without a human decision. Prefer these.
  • HITL — needs a human: a design approval, a balance judgment call, a Studio action only the user can take. UI slices are HITL until an approved UI design artifact exists (see roblox-game-ux); if UI design is missing, create a design-approval slice first and block the UI slice on it.

Process

1. Gather context

Work from the PRD / conversation. If the user passes a reference (issue number, ROADMAP.md line, or file path), read it fully. Connect via roblox-studio-workflow if you need to see current place state. Use the game's own vocabulary.

2. Draft the slices

Order them so the core loop tracer bullet is slice 1, then layer capability outward (persistence → progression → social → monetization → polish). Each later slice should build on a slice that is already playable.

3. Quiz the user

Present the breakdown as a numbered list. For each slice show:

  • Title — short, player-facing.
  • Type — AFK / HITL.
  • Blocked by — which slices must ship first.
  • Playtest check — the exact in-Studio action that proves it works.
  • Server-authoritative? — yes/no and which state it protects.

Ask: is the granularity right (too coarse / too fine)? Are dependencies correct? Should any slice split further? Are HITL/AFK correct? Iterate until approved.

4. Publish

Publish each approved slice to the configured tracker with the needs-triage label/status, blockers first so you can reference real IDs. Mirror each slice as a checkbox in ROADMAP.md using the pack's convention: [ ] not started, [~] built and tested but awaiting approval, [x] only after build + playtest + explicit human approval.

<slice-template>

What to build

The end-to-end player-facing behavior for this slice. Describe what the player does and what the server does — not a layer-by-layer implementation.

Layers touched

  • Client: ...
  • Server (authoritative): ...
  • Remotes: ...
  • Data / persistence: ...
  • UI: ...

Playtest check (acceptance)

  • [ ] In a Studio playtest, <exact action> produces <exact result>.
  • [ ] Valuable state is server-validated and survives rejoin (if applicable).
  • [ ] No new console errors.
  • [ ] Mobile: readable and tappable on a phone-sized viewport (if UI).

UI design artifact

For UI slices: link the approved artifact (see roblox-game-ux) or "blocked on UI design approval".

Blocked by

Reference the blocking slice, or "None — can start immediately". </slice-template>

Do NOT close or modify a parent PRD/issue.

Next

Run `roblox-triage` to sort the slices and mark which are ready to build, then build each slice with its phase specialist (e.g. roblox-luau-architecture, roblox-combat-systems, roblox-ui-implementation, roblox-datastore-persistence) and verify with roblox-playtest-qa. Keep roblox-game-development-lifecycle as the umbrella that decides slice order.

来自同一仓库

更多 Skills

全部 Skills
ashexplained
社区

roblox-playtest-qa

Playtest Roblox experiences through Studio MCP with attention to first-session fun, mobile UX, console errors, progression, social hooks, performance, and exploit-sensitive flows.

1小时安装量
4
GitHub Stars
63
最近更新
7月1日
ashexplained
社区

roblox-animation-audio-feedback

Improve Roblox feel through animation, audio, VFX, haptics-like feedback, reward juice, movement feel, combat feedback, UI transitions, emotes, and satisfying loop feedback without hurting mobile performance.

1小时安装量
3
GitHub Stars
63
最近更新
7月1日
ashexplained
社区

roblox-combat-systems

Design, implement, and review Roblox combat systems. Use for melee, ranged, abilities, cooldowns, hit detection, damage validation, status effects, enemy AI hooks, combat UI, balance knobs, VFX/SFX, and exploit-resistant remotes.

1小时安装量
3
GitHub Stars
63
最近更新
7月1日
ashexplained
社区

roblox-creator-store-security-audit

Audit Roblox Creator Store, Toolbox, inventory, group, universe, package, model, UI, gameplay, plugin-like, and other third-party imported assets for malicious or risky scripts. Use whenever Claude imports, inserts, reviews, or trusts Creator Store assets, especially assets with Scripts, LocalScripts, ModuleScripts, hidden descendants, require calls, HTTP, DataStore, remote creation, obfuscated code, or suspicious names.

1小时安装量
3
GitHub Stars
63
最近更新
7月1日