Evaluation · Wireframes

The six pieces the assessment still needs

Low-fidelity wireframes for the evaluation surfaces named as not-yet-built in assessment-service/README.md. Each maps a cadet's scored footprints from the Starship Academy training world into the enterprise EVE Glyph reference assessment — behind one human-reviewed intake gate.

◇ Wireframe only — structure and flow, not final visual design. No live data, no backend.
Starship Academy wireframe These six evaluation pieces Signed reference assessment
Piece 1 · Submission form / structured intake template

Submit an architecture for assessment

The structured template that opens the WhatsApp routing to the appropriate GitHub author. Routing is the first gate; nothing reaches a repository before a human accepts it.

Intake — routes to author via WhatsApp before any repo update
e.g. NB Power · regulated utility
▾ Regulated utilities / SAP / Healthcare / Marine / Nuclear …
What the system does, where the human-in-the-loop sits, what it integrates…
◯ Yes, public ◉ Private ◯ Decide with author
WhatsApp · LinkedIn · email

Human-reviewed · No self-serve assessment · Author accepts routing before assessment proceeds

Piece 2 · Routing rules engine (operator-matching)

Match the submission to the right author

Operator-matching logic that puts each submission in front of the author whose lattice / education / marketplace / hyperloop work is the relevant precedent. Manual on WhatsApp until routing patterns stabilize.

Routing preview — domain → author
Regulated utilities · SAP lock-indomain match→ EUDG / NB Power author
Iconography proposaldomain match→ Marketplace author
Biological integrationdomain match→ Integration-layer author
Unmatched / novelno author yet→ Lead operator triage

Phase 0: lead operator routes manually. Automation comes when patterns are stable and multiple authors are active.

Piece 3 · Footprint trace format specification

The assessment's time-ordered footprint trace

Every wobble detection, lock-in flag, and re-orientation, in time order — the reproducible record that makes the assessment auditable rather than a black-box verdict.

Footprint trace — sample run (wireframe data)
00:00.0Start-position check — operator reset path presentpass
00:12.4Wobble detected — ingestion faster than operator dampeningwobble
00:19.1Lock-in flag — single-vendor continuity dependencylock-in
00:26.8Re-orientation — human-in-the-loop re-inserted at ingestcorrected
00:31.0Sovereignty reading — 2 exit vectors preservedreading

Format fields (draft): t_offset · glyph · dimension · detail · verdict. Portable (CSV/JSON), forkable, no vendor payload.

Piece 4 · Live-stream surface design

The methodology operating in the open

With submitter and author consent, the assessment streams live — footprints, wobble detections, and lock-in flags shown as they fire, correlated with public-domain event monitoring.

Live-stream layout (consented)
● LIVEAssessment stream · author + submitter
Firing now
▲ Wobble — ingest > dampening
⬡ Lock-in — vendor continuity
◈ Public-domain event correlated: vendor incident, today
↺ Re-orientation accepted by author

Open-governance posture. The stream is the reproducibility record. Off by default; consent required from both sides.

Piece 5 · Public-domain event correlation layer

Read the architecture against the live world

Vendor incidents, AI-safety failures, compliance events, and market signals — the same signals the hyperloop meter/monitor pages track — become inputs to the assessment. An architecture is read against what is happening today.

Vendor incidents

Live feed of lock-in-relevant outages / EOL notices matched to the submitted stack.

Regulated-industry events

Compliance/legislative changes in the submitter's sector, correlated to the sovereignty reading.

AI-safety signals

Public failure-mode events matched to the architecture's safety-reading footprint.

Piece 6 · Public archive of consented assessments

The accumulating corpus of worked examples

With submitter consent, an assessment becomes part of the reference model's public corpus — author-signed, footprint-traced, reproducible.

Public assessment archive (consented only)
Regulated utility — SAP continuity architectureauthor-signedtrace ✓
Healthcare — EHR integration surfaceauthor-signedtrace ✓
Marine — signals telemetry pipelineauthor-signedtrace ✓

The submitter receives a reading, not a verdict. Archival is opt-in. Every entry carries the author attribution and full footprint trace.

Binding · Seven-dimension reading

What every piece feeds into

The cadet's seven test-scores from the Starship Academy world and the submitted architecture produce a quantified reading across the seven assessment dimensions. This is the larger evaluation the wireframes serve.

Quantified reading (wireframe values)
Cognitive fitness (Meth. 13)72
Wobble reading (Meth. 12)48
Lock-in disclosure (Meth. 14)35
Start-position (Meth. 15)80
Monitor maturity (Meth. 16)64
Sovereignty reading58
Safety reading69

A reading, not a verdict. What the enterprise does with it is the enterprise's decision.