Skip to content
SkillModel-led1,391 tokens in use

prototype-fast

Build a throwaway playable prototype in about an hour to answer one question, then decide to keep, kill or retune the mechanic.

9,684 stars on the repository Apache License 2.0gamedev-skills/awesome-gamedev-agent-skills · skills/workflows/prototype-fast
npx skills add gamedev-skills/awesome-gamedev-agent-skills
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing prototype-fast costs

90tokens idle
1,391tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter8/15
  • Structure25/25
  • Examples9/15
  • Output15/15
  • Size15/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity92
  • Maintenance96
  • Community60
  • Maturity24
  • Momentum40

This skill helps a coding agent answer one question about a game idea, usually whether a mechanic is fun. It writes a short brief with a core verb, keep and kill signals and a timebox, then builds a greybox with a debug HUD in a folder the main game never imports.

The user plays it without coaching and reports back, and the agent records a keep, kill or retune decision in the brief. A keep means a rewrite in the real project, not a copy of the prototype code.

What You Get

  • A BRIEF.md template with question, core verb, throwaway flag, timebox, keep signal, kill signal and result
  • A pseudocode shape where only the core mechanic is real and tuning values sit as named constants
  • Written keep, kill and retune rules that rest on observable playtest signals
  • A verify list covering launch, a full loop, the debug HUD, containment in the prototype folder and the recorded result
  • A gotcha list on polishing too early, missing kill criteria and prototypes that grow into the product

Common Use Cases

  • Testing whether a dash or grapple mechanic feels good
  • Deciding whether an idea deserves production time
  • Retuning one timing value and replaying
  • Keeping experiments out of the shipping codebase

What Makes It Unique

The deliverable is a written decision, and the skill requires the kill signal to be fixed before anyone is attached to the code. It asks the user to hand the build to a player with no explanation and report whether the verb was used unprompted within about 30 seconds, so the agent’s own play does not decide the result.

Under The Hood

The agent loads the SKILL.md only; it has no scripts or references. It writes BRIEF.md and the prototype code under a prototypes folder, launches the prototype from the command line or a local web server on localhost, and may drive it with scripted input. It runs git status or searches imports to confirm nothing outside the prototype folder changed. It does not call outside services.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers