Part X. Capstones · Chapter 87
Public Playground MVP
The last capstone turns your strongest original lessons, labs, and diligence tools into one coherent public slice of a learning platform — small enough to finish, honest enough to defend, real enough to teach.
In this chapter 9 sections
Ship one narrow learning loop with a stated input and outcome, semantic explanation, deterministic reference model, accessible interaction, tests, sources, limitations, provenance, and feedback path; add routes only after this slice passes real reader tasks.
A public playground earns trust through one complete learning loop. The Bell-state slice begins with a question, declares basis order, shows state evolution, runs a deterministic reference model, explains measurement statistics, includes a wrong variant that tests catch, and publishes sources, limitations, version, and an issue path. Feature count is irrelevant until that loop works.
Capstone brief: one trustworthy learning loop
The semantic source retains . The accessible explanation says amplitudes occur only on 00 and 11; repeated measurement yields either with equal ideal probability; each shot yields one classical pair. It does not claim faster-than-light signaling or hardware evidence.
The Bell-state slice from question to measurement
Bell’s paper anchors the correlation context. OpenQASM, Qiskit, and Cirq define reference circuit interfaces at the cutoff. ACM’s artifact policy informs release evidence. The full-stack and benchmarking sources require simulator output to remain labeled separately from device performance. GAO supplies public context, not a learning-outcome claim.
Accessibility and provenance are product behavior
from math import sqrt
def evaluate_state(state, tolerance=1e-12):
if type(state) is not list or len(state) != 4 or tolerance <= 0:
raise ValueError("playground state must contain four amplitudes")
if any(type(amplitude) not in (int, float, complex) for amplitude in state):
raise ValueError("amplitudes must be numeric")
norm = sum(abs(amplitude) ** 2 for amplitude in state)
odd_probability = abs(state[1]) ** 2 + abs(state[2]) ** 2
return {"norm": norm, "odd_probability": odd_probability,
"valid": abs(norm - 1.0) <= tolerance and odd_probability <= tolerance}
bell = [1/sqrt(2), 0.0, 0.0, 1/sqrt(2)]
missing_entangler = [1/sqrt(2), 0.0, 1/sqrt(2), 0.0]
unnormalized = [1.0, 0.0, 0.0, 1.0]
reference_result = evaluate_state(bell)
mutation_result = evaluate_state(missing_entangler)
boundary_result = evaluate_state(unnormalized)
invalid_rejected = False
try:
evaluate_state([1.0, 0.0])
except ValueError:
invalid_rejected = True
assert reference_result["valid"] and reference_result["odd_probability"] == 0.0
assert not mutation_result["valid"] and abs(mutation_result["odd_probability"] - 0.5) < 1e-12
assert not boundary_result["valid"] and boundary_result["norm"] == 2.0 and invalid_rejected
print(f"PASS: 87 Bell MVP norm={reference_result['norm']:.1f} mutant_odd={mutation_result['odd_probability']:.1f} boundary_norm={boundary_result['norm']:.1f}")The reference simulator applies a Hadamard to the first qubit and a controlled operation to produce amplitudes [1/sqrt(2), 0, 0, 1/sqrt(2)] in the declared basis. The mutation omits the controlled operation and therefore leaves amplitude on 10 instead of 11. A parity test catches the conceptual failure. The page can explain that result without relying on color or animation.
Artifact contract. A semantic UI contract, deterministic two-qubit simulator fixture, accessibility checklist, release record, and analytic rubric. The fixture preserves normalization and Bell parity; the page works by keyboard and without client-side math; sources, limitations, version, test command, and issue route are visible.
| Release property | Reference record | Acceptance | Mutation |
|---|---|---|---|
| Statevector | [1/sqrt(2),0,0,1/sqrt(2)] | norm and parity | omit entangler |
| Explanation | basis and outcomes in text | no color dependency | hide basis |
| Interaction | keyboard-operable controls | visible focus | remove label |
| Provenance | version, sources, issue route | all visible | stale date |
Exact validation command: python3 tools/validate_briefs.py --briefs data/editorial_briefs_64_87.json --from 64 --through 87 --check-rewritten-sources --execute-artifacts
Release evidence before feature count
Release the slice only after keyboard navigation, focus, 200-percent zoom, narrow layout, no-JavaScript math, deterministic tests, source links, provenance, and issue reporting pass. Record the exact version and expected output. Add a new route only when reader observation shows a distinct question that cannot be answered by improving this one.
Run the public-MVP acceptance suite
Prompt. Build and publish the Bell-state learning slice with a deliberately wrong variant that the tests catch.
Deliverable. Semantic page, accessible state explanation, reference simulator, expected output, failing variant, automated tests, dependency/version record, source ledger, limitations, changelog entry, and feedback route.
Pass condition. The reference test passes, the mutation test fails for the intended reason, keyboard and 200-percent zoom checks pass, and no hardware or learning claim exceeds the evidence shown.
Model repository, launch record, and rubric
Format. Model MVP repository record and 100-point release rubric.
The model repository includes semantic HTML, the four-amplitude fixture, expected output, mutation test, dependency record, source ledger, limitations, changelog, and issue route. Its 100-point rubric allocates 25 to conceptual accuracy, 20 to executable correctness, 15 to accessibility, 15 to evidence and provenance, 15 to pedagogy and assessment, and 10 to release hygiene. The mutation must fail for the intended parity reason.
Verification. The embedded fixture prints the expected Bell probabilities and catches the missing-entangler mutation; the rubric requires accessibility, provenance, and reproducibility for a passing release.
| Criterion | Points | Evidence rule |
|---|---|---|
| Conceptual accuracy | 25 | Show the inspectable conceptual accuracy record; an unsupported assertion receives no credit. |
| Executable reference and mutation | 20 | Show the inspectable executable reference and mutation record; an unsupported assertion receives no credit. |
| Accessibility | 15 | Show the inspectable accessibility record; an unsupported assertion receives no credit. |
| Evidence and provenance | 15 | Show the inspectable evidence and provenance record; an unsupported assertion receives no credit. |
| Pedagogy and assessment | 15 | Show the inspectable pedagogy and assessment record; an unsupported assertion receives no credit. |
| Release hygiene | 10 | Show the inspectable release hygiene record; an unsupported assertion receives no credit. |
| Total | 100 | All pass conditions remain mandatory regardless of point total. |
Companion work
Artifacts for this chapter
These entries resolve to checked-in local source. Commands are reproduced exactly from the chapter manifest, and source-embedded fixtures are exported as direct downloads.
capstone artifact
Bell-state public playground reference slice
Reproduce or test
python3 tools/validate_briefs.py --briefs data/editorial_briefs_64_87.json --from 64 --through 87 --check-rewritten-sources --execute-artifacts
Provenance
Sources and review
- John S. Bell. On the Einstein Podolsky Rosen paradox. Physics Physique Fizika. 1964primary paper
- OpenQASM Technical Steering Committee. OpenQASM 3 specification. Linux Foundation Joint Development Foundation. 2026official technical specification
- IBM Quantum. Qiskit documentation. IBM. 2026official documentation
- Google Quantum AI. Cirq documentation. Google. 2026official documentation
- U.S. Government Accountability Office. Quantum Computing and Communications: Status and Prospects. GAO. 2021government technology assessment
- Association for Computing Machinery. Artifact Review and Badging. ACM Publications. 2026official reproducibility policy
- Lieven M. K. Vandersypen et al.. A look at the full stack. Nature Reviews Physics. 2021peer-reviewed perspective
- Timothy Proctor et al.. Benchmarking quantum computers. Nature Reviews Physics. 2025peer-reviewed perspective
The load-bearing claims in the chapter are mapped inline to this registered source set. A citation supports only the bounded claim beside it.