Skip to main content

Capstone Scoping Lab

This lab is where you turn an idea into a defensible capstone. By the end of a 90-minute session, you will have three written artifacts that unlock the rest of the module.

Retrieval Prompts​

  1. State the three properties of a defensible capstone problem from memory.
  2. What are the four columns of a capstone risk register?
  3. Give the five parts of an MVP definition.
  4. Why is the MVP a vertical slice and not a horizontal one?
  5. What makes a risk "architecturally significant" versus "delivery only"?

Compare and Distinguish​

Separate these pairs clearly:

  • problem statement versus "why this?" memo
  • risk versus TODO versus bug
  • MVP versus prototype versus backlog
  • architecturally-significant risk versus schedule risk
  • "right-sized" versus "cheap"

Common Mistake Check​

For each statement, identify the error and fix it in one sentence:

  1. "My capstone is a URL shortener because I want to use Redis."
  2. "My risk register has 40 rows -- I'm being thorough."
  3. "My MVP is the homepage -- everything else comes later."
  4. "Security is a top-3 characteristic for every system."
  5. "I'll cut tests if I run out of time."

Mini Application: The Three Artifacts​

Produce each of these and commit them to your capstone repo.

Artifact 1 -- Problem statement and "why this?" memo​

Use the concept 01 template literally:

For [user type] who [painful situation], [capstone name] is a [category] that [one-sentence benefit]. Unlike [alternative], our capstone [differentiator]. The capstone works when [observable success condition].

Then a half-page memo answering: who is the user (concretely), what is the painful workflow, what is the alternative, what is the success moment, why is this the right size.

Commit to library/raw/problem.md.

Artifact 2 -- Risk register​

Start with a brain dump (15-30 rows). Collapse and sort. Keep the top 5. For each: risk, mitigation, architecture implication, significance (architecturally significant / delivery only).

Commit to library/raw/risks.md.

Artifact 3 -- MVP definition​

Use the five-part format from concept 03: user, trigger, happy-path steps, success condition, non-goals. Walk through each happy-path step and tag which architectural layer it touches. Write at least 5 non-goals.

Commit to library/raw/mvp.md.

Evidence Check​

This lab is complete only if:

  • Another engineer could read library/raw/problem.md and describe your capstone back to you correctly.
  • Every top-3 risk has both a mitigation and an architecture implication (or is explicitly marked non-architectural).
  • Every happy-path step in library/raw/mvp.md maps to an architectural layer.
  • At least 5 non-goals are written down, and one of them is a feature you actually want to build.
  • You can recite the problem statement, top risk, and MVP success condition from memory in under 60 seconds total.

Integrated Project-Option Scoping​

If you start from a known domain, rewrite it until it is yours.

  1. MERN vertical slice: name the user workflow, API boundary, auth boundary, datastore, and first deploy target.
  2. E-learning system: choose one learner workflow and one instructor workflow; cut everything else.
  3. Restaurant-management system: choose one operational moment, such as reservation arrival through table assignment.
  4. Portfolio-backed capstone: identify the real system behind the portfolio and the evidence the portfolio will expose.
  5. AI assistant integration: write privacy, cost, prompt-boundary, and fallback requirements before implementation.

Evidence check: every option must still produce a vertical slice, risk register, architecture implication, and non-goal list.