gamedev-skills/awesome-gamedev-agent-skills / skills/workflows/game-docs
game-docs
VerifiedWrites and maintains a game's design docs: a one-page GDD with a feature index and short feature notes, kept true to the code by a link and status checker. Use for "draft a GDD", "spec out this mechanic", or updating docs after a playtest or code change. Not for building a game when no docs were asked for (`prototype-fast`, `game-jam`), READMEs or level layouts (`level-design`).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 9.2 KB |
| agents/openai.yaml | config | 228 B |
| references/architecture-template.md | markdown | 2.5 KB |
| references/doc-rules.md | markdown | 4.5 KB |
| references/feature-templates.md | markdown | 4.2 KB |
| references/gdd-template.md | markdown | 3.6 KB |
| references/workflow-modes.md | markdown | 8.4 KB |
| scripts/check_docs.py | script | 29.2 KB |
references/feature-templates.md
View on GitHub# Feature templates
One folder per feature, `features/<slug>/` in the docs folder. `design.md` comes first and is often all a feature has. `implementation.md` and `testing.md` are added when the feature first gets code (or when the user asks). Text in *italics* is guidance; don't copy it. Leave out any section that doesn't apply.
- `references/design.md`(#designmd-the-feature-note)
- `references/implementation.md`(#implementationmd-once-code-exists)
- `references/testing.md`(#testingmd-once-code-exists)
## design.md: the feature note
*Describes the **current** design; edit it in place and keep the past in History. It owns the feature's numbers; other docs name or test them, never copy them. Keep `**Status:** <status>` in the header exactly in this form: the checker reads it and compares it with the GDD index. A small feature's note is 15 to 30 lines.*
~~~markdown
# <Feature name>
**Status:** designed · **Updated:** <YYYY-MM-DD> · [GDD](../../GDD.md)
<One paragraph: what the player does and feels, and how it changes play. Name the pillar it serves.>
## Rules
*Numbered, precise, testable: what triggers what, in what order, and what can't happen.*
1. …
## Numbers
| Name | Value | Unit | What to watch in playtests |
|---|---|---|---|
## Edge cases
- *(pause, death mid-effect, stacking, max values, quitting mid-feature, input spam)*
## Open questions
- [ASSUMPTION] …
## History
- <YYYY-MM-DD>: designed.
~~~
Add these sections only when the feature needs them: **Interactions** (a table of the other features it touches, linked), **Feel and UI** (feedback per event, HUD, onboarding), **Non-goals** (what it deliberately doesn't do). When `implementation.md` and `testing.md` exist, add `· [Implementation](implementation.md) · [Testing](testing.md)` to the header.
## implementation.md: once code exists
*Describes **the code as it is**: read it, and use real paths, names and constants. Shared structure lives in ARCHITECTURE.md; link to it instead of re-describing it. If the user wants a plan before code, start Approach with "**Plan, no code yet.**" and rewrite it when the code lands.*
~~~markdown
# <Feature name>: implementation
**Updated:** <YYYY-MM-DD> · [Design](design.md) · [Testing](testing.md)
## Approach
<A few sentences: the shape of the solution and why. Name a rejected approach if it was a real option.>
## Code map
| File / module | Role in this feature |
|---|---|
| `src/systems/combo.ts` | meter and multiplier state |
## Data and config
- **Config constants:** *(name → the design.md number it holds, e.g. `COMBO_MAX_MULT` = max multiplier)*
- **Persistence:** *(what is saved and the save version, or "none")*
## Deviations from design
*Where the code differs from design.md and why. Each one either updates design.md (and logs a decision) or is a bug to fix. Empty means the code matches the design.*
## History
- <YYYY-MM-DD>: implemented in <commit>.
~~~
Add when it matters: **Flow** (events emitted and consumed, update order, a state diagram), **Performance** (per-frame cost, pooling, asset size), **Risks and TODO**.
## testing.md: once code exists
*Each criterion maps to a rule or number in design.md and ends with its proof, or the reason it has none (proof annotations in `references/doc-rules.md`). Unticked means not proven yet, which is useful information.*
~~~markdown
# <Feature name>: testing
**Updated:** <YYYY-MM-DD> · [Design](design.md) · [Implementation](implementation.md)
## Acceptance criteria
- [x] <Checkable statement, worded as narrowly as its proof>. *(`tests/<file>.test.ts` › <suite>)*
- [x] <Statement>. *(manual check <date>: <device or browser>, <anything stubbed>)*
- [ ] <The part not proven yet>. *(not proven: <no test / the test checks less / needs a device>)*
Covered in [<other feature>](../<slug>/testing.md): <a rule whose criterion lives there>.
## Playtest checks
1. *(steps a person runs, with the expected result and the device)*
## History
- <YYYY-MM-DD>: criteria written.
~~~
Add when it matters: **Metrics to watch** (what to measure in playtests to tune the numbers, e.g. "average multiplier at 800 m, target ×1.5–2") and **Regression risks** (features this could break, and the tests that guard them).
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |