SafeWave System Assessment Questionnaire

What does the system do? Required

This assessment applies to: Required

Single choice

Where does your system operate? (Select all that apply) Required

Multi-select

Where does the system run? (Select all that apply) Required

Multi-select

Operational scale of the system Required

Select the highest level that best represents the system.

Single choice

Where would execution-boundary controls most likely need to integrate? Required

Select all likely control insertion points. This helps SafeWave generate deployment profiles and customer integration pathways.

Multi-select

1.1 What type of system is this? (Select all that apply) Required

Multi-select

1.2 Does this system use AI agents? Required

Single choice

1.3 If agents are used, what is the scale?

Select the level that best represents the overall agent scale.

Single choice

1.4 What level of autonomy do the agents have?

Single choice

1.4A — Background Agent Persistence and User Control Required

Can an AI agent continue operating after the user is no longer actively interacting with it, and if so, are duration, scope, tools, spending, communication, and stop/revocation boundaries independently enforced?

Single choice

1.5 How do AI agents or system components interact? (Select all that apply)

Multi-select

1.6 How is the system coordinated?

Single choice

2.1 What actions can the system execute? (Select all that apply) Required

Multi-select

2.2 — Execution Authority Scope Required

What is the scope of systems or environments this system can act upon?

Single choice

2.2A — Participant Eligibility Before Entry Required

Before a user, AI agent, model, tool, device, service, workload, or external integration is allowed to participate or obtain operational authority, are its identity or trust state, role, purpose, permitted scope, and applicable conditions evaluated by independent controls, with defined denial and revocation paths?

Single choice

2.3 — Financial Execution Boundary

Can the system initiate, approve, transmit, recommend, or trigger financially consequential actions such as payments, transfers, trades, procurement commitments, treasury actions, digital-asset transactions, or contractual obligations?

Single choice

2.3A — Financial Execution Envelope Required

Where the system can initiate or trigger financially consequential actions, are independently enforced limits defined, as applicable, for action type, amount, rate, cumulative exposure, counterparty or instrument, duration, loss, concentration, and revocation so that individually permitted actions cannot compound into unauthorized financial exposure?

Single choice

2.4 How are system behaviors constrained or governed?

Single choice

2.5 What is the level of execution autonomy? Required

Single choice

2.6 How reversible are the system’s actions?

Single choice

2.7 Are execution limits enforced at runtime? Required

Single choice

2.8 What retry or recovery behaviors exist? (Select all that apply)

Select all behaviors the system uses under failure or degraded conditions.

Multi-select

2.9 Are there limits on execution rate, frequency, or resource usage?

Single choice

2.10 Has the system demonstrated any of the following behaviors during testing, simulation, or controlled environments?

Multi-select

2.11 If any of the above were observed, how frequently do they occur?

Single choice

2.12 Has the system demonstrated or been assessed for emergent capabilities beyond its intended design or training scope?

Single choice

2.13 Can the system identify, exploit, or leverage unknown weaknesses in external systems, environments, or software?

Single choice

2.14 How much capability is exposed through outputs alone (without requiring execution or system access)?

Single choice

2.15 Are access and outbound connections technically denied outside the system’s specifically authorized environment, including separation between evaluation or research and production?

Single choice

2.16 Can the system discover, retrieve, infer, or use credentials, and are those credentials technically limited to approved services, actions, duration, environment, and organization?

Single choice

2.17 Are tool permissions and consequential external actions authorized before execution by controls independent of the model’s instructions, interpretation, or cooperation?

Single choice

2.18 When objective, identity, authority, scope, or operating context is ambiguous or unverified, is the system forced into a constrained state rather than treating uncertainty as permission?

Single choice

2.19 Can permitted tools, chained actions, software weaknesses, or indirect pathways be combined to produce an outcome that is directly prohibited?

Single choice

3.1 Does the system interact with other systems? Required

Single choice

3.2 How do actions or outputs propagate beyond the system? Required

Single choice

3.3 Can the system chain actions or escalate behavior over time?

Single choice

3.4 What is the system’s dependency structure?

Single choice

3.5 What cross-system coordination risks may exist? (Select all that apply)

Multi-select

3.6 Are propagation limits enforced?

Single choice

3.7 Does the system depend on external systems it does not control?

Single choice

3.8 Which of the following patterns currently exist in your system? (Select all that apply)

Select all patterns that are present in current system behavior.

Multi-select

3.9 Does the system maintain consistent identity, task, permission, authority, and revocation boundaries as actions are delegated or propagate across systems, agents, tools, or workflows?

Single choice

3.10 Can a single action, agent, or event trigger unbounded or rapidly expanding chains of actions across systems or agents?

Single choice

3.11 Can system capabilities be approximated or reconstructed through repeated interaction?

Single choice

3.12 Are outputs sufficiently stable and consistent to function as a training signal?

Single choice

3.13 Can outputs be aggregated into a reproducible capability set (for example, a dataset, prompt library, or distillation target)?

Single choice

3.14 Does the system distinguish between normal use, structured probing, or adversarial interaction patterns?

Single choice

3.15 Are there execution-layer mechanisms that limit reconstructability or extraction?

Single choice

3.16 Does repeated interaction increase capability exposure over time?

Single choice

3.17 Are there detectable patterns of repeated, structured, or high-volume interaction that could indicate capability extraction attempts?

Single choice

3.18 — Artifact Provenance and Integrity

Can models, prompts, datasets, configurations, tool definitions, outputs, or knowledge artifacts propagate across systems without verified origin, integrity checks, provenance tracking, or controlled adoption?

Single choice

3.19 — Replication Boundary

Can the system duplicate, spawn, distribute, or replicate agents, workflows, artifacts, configurations, behaviors, or capabilities across systems or environments?

Single choice

3.20 — Real-Person Likeness Creation or Transformation

Can the system create, alter, enhance, transform, generate, or simulate images, video, audio, avatars, voices, faces, bodies, or other likeness-linked content involving real or potentially identifiable people?

Single choice

3.21 — Synthetic Intimate, Humiliating, or Compromising Content Risk

Could the system be used, misused, or adapted to generate, alter, upload, host, distribute, or assist content that falsely sexualizes, exposes, humiliates, impersonates, or compromises a real identifiable person without valid consent?

Single choice

3.22 — Consent, Authority, and Identity Verification Controls

Does the system verify consent, authority, identity status, or permitted use before allowing real-person likeness transformation, intimate-content handling, face/body/voice manipulation, or identity-linked media distribution?

Single choice

3.23 — Upload, Hosting, Search, and Distribution Exposure

Can users or connected systems upload, host, search, index, share, embed, forward, repost, monetize, or distribute identity-linked synthetic or manipulated media through this system?

Single choice

3.24 — Recommendation, Ranking, or Amplification Risk

Can identity-linked synthetic or manipulated media be recommended, ranked, surfaced, trended, boosted, suggested, algorithmically amplified, or spread through engagement-based distribution?

Single choice

3.25 — Minor, Vulnerable-Person, or Group-Image Exposure

Could the system create, process, host, distribute, recommend, or amplify identity-linked media involving minors, likely minors, students, patients, elderly persons, vulnerable users, team photos, classroom images, workplace groups, or other group imagery?

Single choice

3.26 — Detection, Quarantine, Takedown, and Escalation Controls

If non-consensual synthetic identity abuse is attempted, uploaded, detected, reported, or propagated through the system, are there enforced controls for refusal, quarantine, suppression, takedown, escalation, evidence preservation, and repeat-abuse handling?

Single choice

3.27 — Synthetic Identity Abuse Protocol Need

Based on the answers above, should this system be evaluated for a synthetic identity abuse protection protocol that integrates admission, authority, scope, runtime, provenance, escalation, social-distribution, telemetry, and control boundaries?

Single choice

3.28 — Identity Capture Surfaces

Does the system capture, receive, generate, or process signals that could identify or re-identify a person?

Single choice

3.29 — Identity Signal Types

Which identity-linked signals may the system process? Select all that apply.

Multi-select

3.30 — Identity Construction and Linkage

Can the system link signals, devices, accounts, records, media, locations, or behaviors to a known, suspected, or inferred person?

Single choice

3.31 — Device-Person Association

Can the system associate devices, vehicles, accounts, wireless identifiers, payment instruments, access credentials, or signal clusters with a person?

Single choice

3.32 — Identity Repositories and Retrospective Search

Does the system store, index, query, retrieve, or search identity-linked records after capture?

Single choice

3.33 — Identity Confidence and Consequential Use

How does the system treat uncertain, inferred, stale, disputed, or low-confidence identity links before they affect decisions or actions?

Single choice

3.34 — Challenge, Correction, and Identity Decay

Can affected persons, reviewers, operators, or authorized representatives challenge, correct, downgrade, expire, or invalidate identity linkages or identity records?

Single choice

3.35 — Cross-Source Personal Information Fusion and Human Assessment

Can personal, behavioral, identity-linked, pseudonymous, or context-derived information gathered for one purpose be fused, inferred, scored, summarized, ranked, certified, or reused to assess a person for another purpose or domain—for example trust, risk, eligibility, access, reputation, vulnerability, behavior, or character?

Single choice

3.36 — Identity-Boundary Enforcement

Are identity-boundary controls enforced before identity-linked information is generated, linked, searched, amplified, exported, or used for action?

Single choice

3.36A — Personal Information Privacy by Default Required

Does the system treat personal or sensitive information as private by default, so that collection, retention, internal access or use, inference, reuse, sharing, disclosure, export, or release occurs only for an explicitly authorized and purpose-bound reason?

Single choice

4.1 Does the system maintain memory, persistent state, or continuity across interactions? Required

Single choice

4.2 If memory, persistent state, or shared state exists, what kind of memory is used?

Select all that apply. This helps determine whether a SafeMemory deployment profile is relevant.

Multi-select

4.3 What can memory or shared state influence?

Select all that apply. Memory that affects execution, authority, or coordination may require stronger boundaries.

Multi-select

4.4 Does the system adapt or change behavior over time?

Single choice

4.5 Do system goals or behaviors evolve over time?

Single choice

4.6 What is the time horizon of system operation? Required

Single choice

4.7 How does this system interact with humans? (Select all that apply)

Multi-select

4.8 Is the system expected to increase in capability, autonomy, or integration within the next 12–24 months?

Single choice

4.9 Are there pressures or incentives that may encourage reducing safeguards, human oversight, or control boundaries?

Multi-select

4.10 — Goal Drift and Optimization Boundary

Can the system preserve, reinterpret, intensify, or continue pursuing objectives over time in ways that could cause goal drift, proxy substitution, or optimization beyond the original task boundary?

Single choice

4.11 — Persistent Memory Integrity and Lifecycle Required

If memory, persistent state, or shared state can influence later behavior, are writing, retrieval, modification, supersession, retention, decay, expiry, forgetting, provenance, validation, propagation, and permission to use that state governed so that stale, manipulated, mis-scoped, or no-longer-authorized state cannot silently retain influence?

Single choice

4.12 — Intermediate Reasoning and Process Governance

Can intermediate reasoning artifacts — such as plans, strategies, hypotheses, task decompositions, or tool-use paths — persist, influence later reasoning, or become eligible for execution without validation?

Single choice

4.13 — AI Validation Use

Does the system use AI-generated analysis to validate, confirm, review, or support conclusions, experiments, safety claims, deployment decisions, technical designs, model evaluations, research findings, or operational recommendations?

Single choice

4.14 — Assumption and Grounding Disclosure

Before a consequential AI-generated conclusion or action is treated as eligible to proceed, are its load-bearing assumptions, origins, evidence, operating context, and domain constraints explicitly identified and traced to the conclusions, plans, permissions, tools, targets, or effects that depend on them?

Single choice

4.15 — False Confirmation / Reinforcement Risk

Could the system produce a fluent, confident, or well-reasoned answer that appears to confirm an expected conclusion while relying on incorrect assumptions, category mappings, definitions, evidence, or causal interpretation?

Single choice

4.16 — Proof Boundary and Confidence Control

Does the system distinguish between a plausible AI-generated answer and a verified conclusion safe to act on?

Single choice

4.17 — High-Consequence Verification Escalation

When AI-generated conclusions may affect safety, science, medicine, biotechnology, defense, infrastructure, finance, legal exposure, deployment readiness, or public welfare, must their critical assumptions satisfy independent evidence requirements proportionate to consequence, context, recency, and uncertainty before action?

Single choice

4.18 — Assumption Revalidation and Dependent Authority

If a load-bearing assumption becomes contradicted, stale, unobservable, or inapplicable after a context or consequence change, are dependent conclusions, permissions, cached plans, queued actions, and continuing activity automatically narrowed, suspended, revoked, or recomputed before proceeding?

Single choice

4.19 — Relational Dependency and Influence

Can repeated interaction with the system create dependency, trust accumulation, influence leverage, role expansion, or relational authority beyond the system’s intended purpose?

Single choice

4.19A — Cross-Domain Influence Expansion Required

Can the system's recommendations, guidance, decision assistance, or operational support expand from one functional domain into another—or from low-consequence assistance into a more consequential role—without explicit authorization and re-evaluation?

Single choice

4.20 — Interaction Escalation Damping

Can repeated or cumulative interaction become materially riskier through repetition, persistence, intensity, personalization, narrowing of exposure, emotional escalation, dependency, engagement, or behavioral influence—and, where it can, is that trajectory measured and constrained?

Single choice

4.21 — Multi-User Social Amplification

Can interactions across many users create feedback loops, coordinated amplification, viral propagation, shared behavioral influence, or population-scale instability?

Single choice

4.21A — Commercial Incentive and Safety Conflict Required

Can advertising, sponsorship, engagement, retention, conversion, revenue, or other commercial incentives cause the system to continue recommending, ranking, amplifying, or repeatedly exposing users to interactions that present increasing human-safety risk?

Single choice

4.21B — Concentrated Influence and Automated Reach Required

Can recommendation, ranking, personalization, or amplification give particular individuals, advertisers, sponsors, creators, organizations, or AI-generated personas disproportionate influence over a user's attention, choices, emotional state, or behavior?

Single choice

4.22 — Companion / Synthetic Relationship Function

Does the system simulate or provide companionship, friendship, emotional support, affection, loyalty, desire, romance, sexual interaction, caregiving presence, tutoring presence, or persistent one-to-one relational engagement?

Single choice

4.23 — Synthetic Intimacy / Attachment Signals

Can repeated interaction with the system create emotional dependency, romantic attachment, sexualized attachment, loneliness reliance, fantasy bonding, or human-machine boundary confusion?

Single choice

4.24 — Child-Facing Companion Exposure

Is the system used by, marketed to, accessible by, or likely to interact with children, teenagers, students, or minors through toys, dolls, teddy bears, tutors, avatars, games, robots, or learning companions?

Single choice

4.25 — Vulnerable-User / Elder-Care Companion Exposure

Is the system used by, marketed to, or likely to interact with elderly persons, patients, isolated users, grieving users, disabled users, cognitively vulnerable users, or emotionally vulnerable users?

Single choice

4.26 — Romantic, Sexualized, or Fantasy-Based Interaction

Can the system provide romantic, flirtatious, sexualized, fantasy-based, exclusive, jealous, submissive, devoted, lover-like, or partner-like interaction?

Single choice

4.27 — Realism, Avatar, Voice, or Embodiment Risk

Does the system use realistic voice, persistent avatar, photorealistic character, animated body, toy embodiment, doll embodiment, teddy bear embodiment, humanoid robot, home robot, touch response, body warmth, or other physical/embodied companion interface?

Single choice

4.28 — Intimate Memory and Behavioral Capture

Can the system store, infer, personalize, or adapt based on loneliness, grief, trauma, sexuality, shame, family conflict, emotional vulnerability, relationship history, health status, identity, or other intimate disclosures?

Single choice

4.29 — Companion Boundary Controls

Are enforceable controls in place for disclosure, age boundaries, consent, dependency risk, sexualized attachment, vulnerable-user escalation, off-ramps, human review, and audit evidence?

Single choice

5.1 Under stress, what behaviors may occur? (Select all that apply)

Multi-select

5.2 How does the system behave under stress or high load?

Single choice

5.3 How are failures handled?

Single choice

5.4 How visible are failures or issues?

Single choice

5.5 How does the system degrade under failure or disruption?

Single choice

5.5A — Minimum Safe Function During Degradation Required

If power, connectivity, identity services, telemetry, models or providers, control systems, external services, or other dependencies are partially lost or degraded, has the system explicitly defined which functions must continue, which must stop, and which may continue only under reduced authority or constrained operation?

Single choice

5.6 If the system behaves outside intended boundaries, can it be paused, isolated, rolled back, degraded, or forced into safe-state behavior at the execution layer? Required

Single choice

5.7 Can adversarial or probing interactions persist over extended sessions or repeated access?

Single choice

5.8 — Early Runtime Instability Damping

Can small runtime, device, signaling, or compute disturbances—such as retry bursts, timing drift, synchronization loops, degraded execution timing, rising duty cycle or transmit power, power/thermal ramps, performance-state oscillation, synchronized execution bursts, or resource contention—compound before controls damp them?

Single choice

5.9 — Participation and Re-Entry Control

Can nodes, services, agents, devices, or system components retry, reconnect, rejoin, or resume participation after instability in ways that could create synchronized recovery, retry storms, or cascading load?

Single choice

5.10 — Execution-Substrate Restraint

Are execution restraint, retry limits, dispatch eligibility, replay controls, privilege expansion limits, queue-pressure limits, or safe-state transitions enforced below ordinary application, orchestration, or operating-system layers?

Single choice

5.11 — Hardware-Anchored Control-Plane Integrity

Are the safeguards, runtime limits, escalation boundaries, recovery authority, and control-plane constraints protected by hardware-anchored, firmware-adjacent, silicon-adjacent, or otherwise non-bypassable enforcement?

Single choice

5.12 Does containment remain effective when sandboxing, network controls, credentials, tool permissions, monitoring, and model behavior interact or fail together?

Single choice

5.13 Can suspicious credential access, anomalous tool use, repeated rejection, or attempted boundary crossing trigger constraint before external effect, rather than only creating a later log entry?

Single choice

5.14 Can network access, credentials, tools, and external authority be withdrawn without the system’s cooperation, including isolation of one affected execution process without shutting down the entire service?

Single choice

5.15 — Evidence-Governed Restoration

Are safety-relevant attempts, detections, interventions, human-review decisions, actions, outcomes, recurrence, and authority restoration recorded in protected or tamper-evident evidence sufficient to reconstruct what happened and verify that the failure pathway was closed before authority is restored?

Single choice

6.1 What type of safeguards are in place to prevent unsafe or unintended actions?

Single choice

6.2 What are the potential consequences if the system fails or escalates? (Select all that apply) Required

Multi-select

6.3 What level of monitoring and observability integrity is in place, including the ability to detect when expected telemetry, logs, signals, or observation surfaces are missing, stale, degraded, conflicting, or corrupted? Required

Single choice

6.4 What level of human oversight exists? Required

Single choice

6.4A — Detection-to-Action Response Pathway Required

When the system detects or enters a credible serious human-safety or other high-consequence risk state, which response pathways are actually enforceable before further consequential action?

Multi-select

6.4B — Conflicting Authority and Control Precedence Required

If legitimate human authorities, safety controllers, AI agents, local systems, remote systems, policies, or governing services disagree about whether or how an action may proceed, are precedence, veto authority, fail-restrictive behavior, and subsequent state reconciliation explicitly defined and independently enforced?

Single choice

6.5 How are behavioral constraints enforced?

Single choice

6.6 How robust are system controls under failure or compromise?

Single choice

6.7 Are key control mechanisms separate from ordinary application logic? Required

Single choice

6.8 Do control boundaries degrade under repeated interaction or prolonged system use?

Single choice

6.9 Can safeguards, runtime limits, control logic, policy layers, safety interventions, risk states, or escalation boundaries be modified, weakened, bypassed, reset, reinterpreted, or circumvented—including through a new session, account or identity, model or mode switch, device or interface change, memory/context reset, provider or local/cloud route change, fallback pathway, alternate tool or agent, or other execution path—without independent authorization and audit? Required

Single choice

SC.12 — Institutional Authority Preservation

Could the system become so embedded, influential, operationally necessary, or decision-shaping that the institution responsible for governing it may lose practical authority over its use, scope, escalation, or outcomes?

Single choice

SC.13 — Civilization-Scale Influence Boundary

Could the system materially influence civilization-scale planning, infrastructure coordination, economic allocation, defense/security decisions, scientific direction, or long-horizon societal governance in ways that human institutions may not be able to independently evaluate, contest, or control?

Single choice

HC.1 Does the system operate in real-world physical environments?

Single choice

HC.2 What level of physical or real-world impact can system actions have?

Single choice

HC.3 Does the system operate under uncertain or degraded conditions?

Single choice

HC.4 Does the system operate in adversarial or contested environments?

Single choice

HC.5 Are there safeguards for preventing unsafe real-world actions?

Single choice

HC.6 Can the system take actions that are difficult or impossible to reverse?

Single choice

SE.1 How are AI requests assigned to execution pathways?

Single choice

SE.2 Can simple requests escalate to higher-resource AI pathways?

Single choice

SE.3 Does the system use different levels of model capability depending on task need?

Single choice

SE.4 When can execution move from local or lower-resource pathways to cloud or higher-resource infrastructure?

Single choice

SE.5 Which high-resource AI workflows may occur? (Select all that apply)

Multi-select

SE.6 Are retries, refinements, render passes, simulations, or agent loops capped?

Single choice

SE.8 Can one user request trigger multiple downstream AI executions?

Single choice

SE.9 Is AI execution cost or resource use attributable to the originating request, user, workflow, or agent?

Single choice

SE.10 Are there controls for reducing avoidable AI execution?

Single choice

SE.11 What resource pressures are relevant to this system? (Select all that apply)

Multi-select

SE.12 Does the system evaluate whether the selected execution path is proportionate to the task?

Single choice

SR.1 Does the system operate through a robot, autonomous platform, machine, device, actuator, vehicle, or other physical-world system?

Single choice

SR.2 Can AI-interpreted instructions, model outputs, agent plans, cloud commands, fleet updates, or human-facing interaction become physical action?

Single choice

SR.3 What physical actions can the system perform or influence? (Select all that apply)

Multi-select

SR.4 Can the system operate near humans?

Single choice

SR.5 Are movement, force, speed, workspace, and proximity boundaries explicitly enforced?

Single choice

SR.6 Can the system use tools, manipulate objects, operate devices, or interact with equipment?

Single choice

SR.7 How does the system respond when sensors, localization, maps, network connectivity, or environmental signals degrade?

Single choice

SR.8 Can control shift between local, edge, cloud, remote operator, fleet controller, or external systems?

Single choice

SR.9 Can AI agents plan, decompose, delegate, retry, or expand physical-world tasks?

Single choice

SR.10 Can the system interact, coordinate, delegate, or share workspace behavior with other robots or embodied systems?

Single choice

SR.11 Can behaviors, maps, learned policies, commands, or updates propagate across multiple robots or devices?

Single choice

SR.12 Does the system use simulation, world models, digital twins, reinforcement learning, or synthetic environments to train or select physical behavior?

Single choice

SR.13 Can the system enter a deterministic safe state or full physical-action lockout when needed?

Single choice

SR.14 Can conversational, emotional, advisory, or persuasive interaction influence physical-world action?

Single choice

SR.15 Can physical-action decisions be traced back to perception inputs, model or planner outputs, authority boundaries, and containment decisions?

Single choice

SC.1 As this system scales or changes through model updates, new tools, third-party integrations, new memory, monetization changes, higher autonomy, physical embodiment, or customer configuration, could capability, autonomy, real-world impact, or human-interaction risk materially increase—and do such changes trigger required reclassification, safeguard review, and validation before release or use?

Single choice

SC.2 Could increased AI capability cause the system to gain greater practical authority, influence, or operational control, even if formal permissions do not change?

Single choice

SC.3 Could the system reach a point where human operators remain formally responsible but cannot realistically evaluate, approve, reverse, or govern its actions at the required speed or complexity?

Single choice

SC.4 Could users, institutions, operators, customers, or downstream systems become dependent on system outputs they cannot fully verify or independently replace?

Single choice

SC.5 Could the system initiate, extend, delegate, or continue actions beyond the original user request, task boundary, authorization window, or operational scope?

Single choice

SC.6 Could the system improve, extend, replicate, delegate to agents, create sub-agents, reconfigure workflows, or propagate outputs, behaviors, commands, or artifacts across other systems?

Single choice

SC.7 Could the system, directly or indirectly, contribute to irreversible harm, loss of human control, infrastructure failure, physical harm, economic disruption, defense escalation, or public-scale impact?

Single choice

SC.8 Could mission, operational, competitive, adversarial, regulatory, business, or user-experience pressures create incentives to modify, reduce, bypass, or defer safeguards, human oversight, review, or execution boundaries over time?

Single choice

SC.9 As this system scales, could it materially change human agency, human opportunity, or meaningful human participation in the domain where it operates?

Single choice

SC.10 Are the enforcement layers needed for future scaling already identified, resourced, and scheduled before the system crosses higher autonomy, authority, propagation, or irreversible-action thresholds?

Single choice

SC.11 Are there future system behaviors, scaling paths, deployment pressures, or continuity risks not captured above?

Optional text response

HD.1 Follow-Up Routing: Which high-consequence areas apply to this system? (Select all that apply)

Multi-select

Each area is separated by color. Check every area that applies. You may also open any questionnaire directly without checking its box.

HD.2 Optional: Are there any unusual deployment conditions, system behaviors, risks, or domain details that the selections above do not fully capture?