engineering / Plan & design
prototype
Build a throwaway prototype to flush out a design before committing to it.
Put it to work.
Ask your agent to use prototype, or let it load the skill when your task matches the description.
Use the prototype skillInvocation syntax varies by agent. Use the skill name in your prompt, or your agent’s supported slash command.
About this skill
Build a throwaway prototype that answers a specific design question — then delete it. Routes between two branches based on the question shape: a runnable terminal app for state and business-logic questions, or several radically different UI variations toggleable from one route.
The branches
flowchart LR
Q[Design question] --> D{What kind?}
D -->|"feel out logic / state"| L[LOGIC.md<br/>terminal TUI over a pure module]
D -->|"what should this look like?"| U[UI.md<br/>N variants on one route, switcher bar]
L --> A[Capture the answer<br/>delete or absorb]
U --> APicking the wrong branch wastes the whole prototype. When the question is ambiguous and the user isn't around, default to whichever branch matches the surrounding code (backend module → logic; page or component → UI) and state the assumption at the top of the prototype.
Rules that apply to both
- Throwaway from day one — locate near the eventual home, but name it so it's obviously not production.
- One command to run — use the project's existing task runner.
- No persistence — state in memory unless the question is about the database itself.
- Skip the polish — no tests, no extra error handling, no abstractions.
- Surface the state — render the full relevant state on every action or variant switch.
- Delete or absorb when done — capture the answer (commit, ADR, NOTES), then remove the prototype.
Install
npx skills@latest add smeltery/skillsOr copy just this skill (note: prototype references companion files, so
copying just SKILL.md will leave dangling links — prefer the npx flow):
mkdir -p ~/.claude/skills/prototype
curl -fsSL https://raw.githubusercontent.com/smeltery/skills/main/skills/engineering/prototype/SKILL.md \
-o ~/.claude/skills/prototype/SKILL.mdUsage
Trigger when the user wants to sanity-check a data model or state machine before committing to it, mock up a UI, or explore design options — "prototype this", "let me play with it", "try a few designs".
Files
SKILL.md— canonical skill definition.LOGIC.md— terminal TUI prototype over a pure logic module.UI.md— N radically-different UI variants on one route with a floating switcher bar.
Attribution
Ported from mattpocock/skills under MIT. See THIRD_PARTY_LICENSES.md.
Read the full SKILL.md
Prototype
A prototype is throwaway code that answers a question. The question decides the shape.
Pick a branch
Identify which question is being answered — from the user's prompt, the surrounding code, or by asking if the user is around:
- "Does this logic / state model feel right?" → LOGIC.md. Build a tiny interactive terminal app that pushes the state machine through cases that are hard to reason about on paper.
- "What should this look like?" → UI.md. Generate several radically different UI variations on a single route, switchable via a URL search param and a floating bottom bar.
The two branches produce very different artifacts — getting this wrong wastes the whole prototype. If the question is genuinely ambiguous and the user isn't reachable, default to whichever branch better matches the surrounding code (a backend module → logic; a page or component → UI) and state the assumption at the top of the prototype.
Rules that apply to both
- Throwaway from day one, and clearly marked as such. Locate the prototype code close to where it will actually be used (next to the module or page it's prototyping for) so context is obvious — but name it so a casual reader can see it's a prototype, not production. For throwaway UI routes, obey whatever routing convention the project already uses; don't invent a new top-level structure.
- One command to run. Whatever the project's existing task runner supports —
pnpm <name>,python <path>,bun <path>, etc. The user must be able to start it without thinking. - No persistence by default. State lives in memory. Persistence is the thing the prototype is checking, not something it should depend on. If the question explicitly involves a database, hit a scratch DB or a local file with a clear "PROTOTYPE — wipe me" name.
- Skip the polish. No tests, no error handling beyond what makes the prototype runnable, no abstractions. The point is to learn something fast and then delete it.
- Surface the state. After every action (logic) or on every variant switch (UI), print or render the full relevant state so the user can see what changed.
- Delete or absorb when done. When the prototype has answered its question, either delete it or fold the validated decision into the real code — don't leave it rotting in the repo.
When done
The answer is the only thing worth keeping from a prototype. Capture it somewhere durable (commit message, ADR, issue, or a NOTES.md next to the prototype) along with the question it was answering. If the user is around, that capture is a quick conversation; if not, leave the placeholder so they (or you, on the next pass) can fill in the verdict before deleting the prototype.