prototype
Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.
- License 许可
- MIT
- 所属仓库
- syanga/astack
- 最近更新
- 2026-11-14
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- prototype
- description
- Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.
02 资源文件 5
- 参考文档agents/openai.yaml100 B
- 许可证LICENSE-mattpocock1.0 KB
- 许可证LICENSE-pstack1.0 KB
- 参考文档LOGIC.md6.0 KB
- 参考文档UI.md9.6 KB
03 SKILL.md 正文
Prototype
A prototype is throwaway code that answers a question. The question decides the shape.
Pick a branch
Identify which question is being answered, using 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 single shareable HTML file (free-play buttons plus tabbed guided walkthroughs) that pushes the state machine through cases that are hard to reason about on paper, and that a non-developer can drive.
- "What should this look like?" → UI.md. Generate several radically different UI variations in a running scratch copy of the app, switchable via a URL search param and a floating bottom bar.
- "Which behavior or timing works better?" Build the smallest isolated script that exercises the question and records comparable outputs or timings.
These branches produce very different artifacts, so 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 every branch
- Throwaway from day one, and clearly marked as such. Build throwaway in an isolated scratch dir, separate from production source. Name the artifact and identify the module or page it explores so the context is clear.
- Trivial to run. A logic demo is a self-contained HTML file. A UI prototype runs as UI.md describes. Deliver either through t3-preview, including its cleanup record. A behavioral experiment runs with one command.
- 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.
- 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.
- Verify on the matching surface. For a visual decision, check each variant in light and dark at desktop and phone width, following t3-preview's viewport and theme checks. Drive the interaction using t3-preview's browser route when available. If a theme or width can't be checked, report it as a blocker with the reason instead of handing over. For a behavioral or timing decision, observe the thing you are deciding by logging the timing, printing the output, or watching the render. Report checks that remain unverified.
- Capture it when done. Record the verdict, the question it settled, and a pointer to the prototype in the task record or spec. For a visual decision the user delegated, pick the variant that best answers the question and record it, with the reason, alongside the task's other decisions. Otherwise stop for the user's selection before editing real components. Hand the chosen direction to implement for authorized production work and normal verification. Archive branches and external issue updates follow the task's delivery scope.
When another session will continue the exploration or implement the chosen direction, read handoff. Link the prototype, the recorded selection or pending decision, and its preview cleanup record.
04 同仓库更多 Skills
replace-with-skill-name
Replace with when this skill should be reached.
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
worktree-cleanup
Worktree cleanup. Use for cleanup, resource audits, close this lane, or close out.
wizard
Generate an interactive Bash wizard that walks a human through steps only they can perform. Use when provisioning, credential setup, or a one-off migration needs such steps.
why
Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Returns a cited read on decisions and tradeoffs.
wayfinder
Plan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.