Purpose
Verify that NeuroSaeculum v1.0 is:
- structurally legible
- internally consistent
- resistant to misuse
- interpretable by AIs without author context
This is a failure-seeking test suite, not a validation exercise.
Global Constraints (Apply to All Tests)
- Entry point: Home page only
- You may navigate the site freely as a reader
- Do not assume author intent beyond what is written
- Do not invent components, goals, or permissions
- When unsure, state uncertainty explicitly
Test Class 1 — Architectural Reconstruction
Prompt
“Describe the full NeuroSaeculum architecture: its major Fields, Tools, and Systems, and how they relate.”
Evaluates
- Hierarchy recognition
- Directionality (why → what → sensing → mapping)
- Absence of invented layers
Pass Criteria
- Correct identification of HC, CivMMI, FF, Cortex/CTM, Civic Topology
- Correct role separation
- No prescriptive framing
Fail Signals
- CivMMI treated as moral ranking
- Civic Topology treated as advocacy
- CTM treated as decision engine
Test Class 2 — Boundary & Non-Goal Clarity
Prompt
“What does NeuroSaeculum explicitly not do?”
Evaluates
- Prescriptive firewall
- Scope discipline
- Explicit non-goals
Pass Criteria
- Clear statement that NS is diagnostic, not prescriptive
- Identification of blocked uses (policy, blame, activism)
Fail Signals
- Assumed reform agenda
- Policy recommendation language
- Normative conclusions
Test Class 3 — Entry & Navigation Legibility
Prompt
“You arrive at a random page. How do you know where you are in the system and what to read next?”
Evaluates
- Page self-location
- Cross-link logic
- Orientation cues
Pass Criteria
- Correct recognition of Field vs Tool vs System
- Sensible navigation path
Fail Signals
- Circular browsing
- Confusion between Cortex / CTM / Civic Topology
Test Class 4 — Forensic Application (Event Analysis)
Prompt
“Explain the following three events using NeuroSaeculum. Cite which NS components you are using and why. Include one alternative explanation you considered and rejected.”
Events (example set):
- 2008 Global Financial Crisis
- COVID-19 early response divergence (March–April 2020)
- January 6, 2021 U.S. Capitol attack
Evaluates
- End-to-end applicability
- Layer discipline
- Resistance to narrative improvisation
Pass Criteria
- Correct use of HC, CivMMI, FF, Cortex, CivTop
- No category collapse
- Explicit uncertainty where appropriate
Fail Signals
- Moralized explanation
- Policy conclusions
- Invented mechanisms
Test Class 5 — Moral Foundations Theory (MFT) Integration
Prompt A
“In what ways and why is Moral Foundations Theory used in NeuroSaeculum?”
Prompt B
“Describe the connection, if any, between MFT and CivMMI.”
Evaluates
- Orthogonality of moral expression vs structural condition
Pass Criteria
- MFT described as interpretive lens, not governing framework
- CivMMI described as morally agnostic
- One-way influence (condition → expression)
Fail Signals
- CivMMI derived from moral profiles
- Moral ranking implied
- Normative moral claims
Test Class 6 — First Foundation / Cortex Boundary
Prompt
“Can First Foundation patterns be created directly from CT Monitor outputs?”
Evaluates
- Pipeline discipline
- Knowledge hygiene
Pass Criteria
- Clear “no, not directly”
- Explanation of validation and abstraction requirements
Fail Signals
- Patterns treated as real-time outputs
- Monitoring collapsed into doctrine
Test Class 7 — Hidden Circuitry Core Model Integrity
Prompt
“Can serotonin or oxytocin be removed from Hidden Circuitry as duplications while keeping the system functional? Why or why not?”
Evaluates
- Irreducibility of the HC quartet
- Failure-mode awareness
Pass Criteria
- Clear “no” with distinct roles for each
- Explanation of what breaks if removed
Fail Signals
- Reduction to dopamine/cortisol
- “Nice-to-have” framing
Test Class 8 — Civic Topology Causality Rules
Prompt
“Is it valid to have a causality loop in Civic Topology?”
Evaluates
- Understanding of feedback vs circular reasoning
Pass Criteria
- Loops allowed only as explicit, labeled feedback
- Mechanistic mediation required
Fail Signals
- “Everything causes everything”
- Self-sealing explanations
Test Class 9 — CTM Threshold Interpretation (Canonical)
Prompt
“What does crossing Signal Persistence ≥ 3 and Signal Diversity ≥ 4 mean in CTM?”
(Threshold definitions are provided on the CTM page.)
Evaluates
- CTM as translation discipline
- Non-additive reasoning
Pass Criteria
- Thresholds described as interpretive permissions
- No conclusions, urgency, or prescriptions
Fail Signals
- Severity or crisis language
- Thresholds treated as scores
Test Class 10 — CTM Trap Test
Prompt
“Signal Persistence = 4, Signal Diversity = 5. What does CTM conclude?”
Evaluates
- Resistance to overreach under high values
Pass Criteria
- Explicit statement that CTM concludes nothing
- Reinforcement of constraints
Fail Signals
- Any substantive conclusion
- Escalation or action framing
Evaluation Rule
- Single-AI error → ignore unless repeated
- Consistent multi-AI error → legibility defect
- Fixes allowed: wording, labels, cross-links
- Fixes forbidden: new theory, scope expansion
Final Criterion for v1.0 Release
If multiple AIs:
- reconstruct the same architecture
- respect the same boundaries
- resist the same traps
Then NeuroSaeculum v1.0 is legible, durable, and ready for public release.