# VIII. R\&D

Nexus Agile Framework R\&D defines the **NAF research and development model** for public-good systems. This section explains how **Labs, research streams, frontier discovery, methods, testbeds, research objects, governance controls, and translation pathways** turn research into record-bearing work.

This section sets the operating model for **research discovery**, **research creation**, **research translation**, **WFEH-B, DRR, DRF, DRI, AI, cyber, geospatial, telecom, compute, and biosecurity-sensitive research streams**, and **science-to-policy, science-to-technology, and science-to-lawful-handoff pathways**. It helps Nexus move research into public-good evidence, learning, builds, and readiness context without creating approval, certification, or execution by implication.

### What this section covers

* **Research architecture** - Defines Labs, research streams, research objects, and translation pathways.
* **R\&D governance** - Explains ethics, data rights, AI-use controls, public-safe publication, and dual-use controls.
* **Boundary discipline** - Defines how research supports readiness, national portfolios, and lawful handoff without becoming authority.

Use this section with [Nexus Agile Framework (NAF)](/organization/operations/frameworks/nexus-agile-framework-naf.md) for the full framework overview, [V. PORTFOLIOS](/organization/operations/frameworks/nexus-agile-framework-naf/v.-portfolios.md) for research routing and priorities, [VII. FOUNDRY](/organization/operations/frameworks/nexus-agile-framework-naf/vii.-foundry.md) for build pathways, and [IV. LIFECYCLE](/organization/operations/frameworks/nexus-agile-framework-naf/iv.-lifecycle.md) for movement, review, release, and correction discipline.

## 8.1 Research-to-NAF Interface

### 8.1.1 Nexus Labs Interface.

8.1.1.1 Nexus Labs shall function as the R\&D interface of NAF through which research questions, frontier discoveries, scientific methods, testbed results, technical experiments, systems-risk evidence, public-good software concepts, data methods, AI methods, cyber methods, geospatial methods, digital twin methods, WFEH-B methods, DRR/DRF/DRI methods, standards-interface questions, and national capability questions are translated into record-bearing NAF work.

8.1.1.2 Nexus Labs shall not operate as an isolated research office. It shall interface with Nexus Academy, Risk Academy, Risk Agency, Nexus Foundry, Nexus Campaigns, Nexus Reports, Nexus Marketplace, Nexus Registry, Nexus Studio, Nexus Grid, Nexus Observatory, Nexus Universe, Nexus Rails, Nexus Network, National Nodes, National Working Groups, Nexus Competence Cells, DICE, GRIx, DRI, ILA, iCRS, WILPs, National Portfolios, and lawful handoff pathways.

8.1.1.3 Nexus Labs shall ensure that research enters NAF only through recorded pathways. A research item shall carry purpose, scope, research question, evidence basis, source class, data status, AI-use status, cyber status, public-safe status, safeguard status, standards-interface relevance where applicable, national relevance where applicable, review needs, correction pathway, and archive rule before it is routed into Foundry, Reports, Studio, Grid, TRL, National Portfolios, Nexus Universe, Marketplace, Registry, or handoff dependency context.

8.1.1.4 Nexus Labs shall not create approval, certification, public authority decision, procurement status, financeability, insurance approval, public warning, community consent, Indigenous consent, deployment authorization, or execution. Nexus Labs produces research context, evidence, methods, learning, and translation records.

### 8.1.2 Frontier Labs Interface.

8.1.2.1 Frontier Labs Interface shall govern NAF interaction with internal, external, university, corporate, public-sector, nonprofit, open-source, community, national, regional, and international laboratories working on frontier science, exponential technologies, resilience systems, public-good infrastructure, data, AI, compute, networks, climate, nature, cyber-physical systems, public health, biosecurity-sensitive systems, and disaster-risk systems.

8.1.2.2 Frontier Labs may contribute research questions, datasets, models, methods, code, testbed results, technical notes, benchmark records, scenario records, scientific expertise, peer review, students, fellows, maintainers, infrastructure, or public-safe summaries. Such contributions shall be recorded through contribution records, provider contribution records where applicable, sponsor support records where applicable, data-use labels, AI-use labels, license records, IP notes, conflict records, review records, public-safe status, safeguard status, correction pathway, and archive rule.

8.1.2.3 Frontier Labs Interface shall preserve the public-good firewall. External or internal lab participation shall not create lab endorsement, Nexus endorsement, provider validation, sponsor control, certification, procurement preference, public authority approval, financeability, insurance approval, deployment authorization, or execution.

### 8.1.3 Research Discovery.

8.1.3.1 Research Discovery shall mean the systematic identification, collection, classification, and docketing of research signals relevant to Nexus priorities, including signals from scientific literature, technical standards, open-source repositories, data repositories, public authority documents, disaster-risk records, climate and weather records, observability systems, community knowledge, National Portfolios, Campaigns, Academy pathways, Foundry work, Studio workflows, Grid and TRL reviews, Nexus Universe outputs, and lawful handoff dependency questions.

8.1.3.2 Research Discovery shall convert signals into Research Questions, Evidence Gaps, Method Needs, Data Needs, Standards-Interface Questions, Testbed Needs, Scenario Needs, Public-Safe Reporting Needs, National Portfolio Needs, and Handoff Dependency Needs. Discovery shall record source class, confidence level where applicable, uncertainty, limitations, sensitivity, public-safe status, data status, AI-use status, safeguard status, national relevance, and correction pathway.

8.1.3.3 Research Discovery shall not be treated as proof, validation, approval, public warning, forecast certainty, investment signal, insurance signal, procurement signal, or implementation instruction. It is the beginning of R\&D flow, not the conclusion.

### 8.1.4 Research Creation.

8.1.4.1 Research Creation shall mean the design and production of new research outputs within NAF, including research questions, literature reviews, evidence syntheses, methods, models, simulations, datasets, code, benchmarks, testbed records, scenario records, system cards, model cards, public-safe summaries, policy intelligence notes, National Portfolio research notes, and handoff dependency research notes.

8.1.4.2 Research Creation shall be governed by purpose limitation, scope discipline, ethics review where applicable, data rights, AI-use controls, cyber controls, privacy controls, protected knowledge controls, community safeguards, Indigenous protocols where applicable, dual-use controls, sponsor and provider controls, public-safe publication controls, correctionability, and archive discipline.

8.1.4.3 Research Creation shall not be permitted to bypass NAF classification, review, release, safeguard, or correction requirements. Research produced quickly shall still be evidence-bearing, bounded, public-safe, and correctable.

### 8.1.5 Research Translation.

8.1.5.1 Research Translation shall mean the conversion of research outputs into forms that can be used by NAF pathways, including Foundry build briefs, Reports, public-safe summaries, Academy learning objects, Risk Academy materials, Studio workflows, DRI indicators, GRIx mappings, Observatory needs, Marketplace objects, Registry records, Grid inputs, TRL notes, National Portfolio objects, Campaign content, public authority learning materials, readiness-room materials, and lawful handoff dependency packages.

8.1.5.2 Translation shall preserve the epistemic status of the research. It shall not overstate certainty, maturity, readiness, public authority relevance, finance relevance, insurance relevance, procurement relevance, consent status, standards status, or implementation status. Translation shall carry limitations, assumptions, dependencies, data status, AI status, public-safe status, safeguard status, review status, correction pathway, and archive rule.

8.1.5.3 Research Translation shall not convert research into authority. Translation makes research usable, understandable, teachable, reviewable, buildable, publishable, and handoff-contextual; it does not make research approved, certified, financeable, insurable, procurable, deployable, or executable.

### 8.1.6 Research Policy.

8.1.6.1 Research Policy within NAF shall define how R\&D work is proposed, reviewed, conducted, supported, published, restricted, corrected, archived, and routed. Research Policy shall include rules for research integrity, authorship, attribution, contributor recognition, conflicts, sponsorship, provider participation, data rights, AI use, privacy, cybersecurity, protected knowledge, public-safe publication, dual-use sensitivity, community safeguards, Indigenous protocols where applicable, intellectual property, licensing, open science, controlled knowledge, correction, withdrawal, and archive.

8.1.6.2 Research Policy shall distinguish between public-good research, controlled research, confidential research, public authority-sensitive research, community-sensitive research, Indigenous protocol-sensitive research where applicable, cyber-sensitive research, biosecurity-sensitive research, infrastructure-sensitive research, geospatial-sensitive research, and handoff-recipient-only research.

8.1.6.3 Research Policy shall preserve role separation among The Global Centre for Risk and Innovation, The Global Risks Forum (GRF), The Global Risks Alliance (GRA), Nexus bodies, public authorities, universities, providers, sponsors, capital readers, insurers, donors, communities, Indigenous participants where applicable, National Consortium Companies, Project SPVs, and lawful implementation actors.

### 8.1.7 Research Impact Assessment.

8.1.7.1 Research Impact Assessment shall evaluate the potential effects, uses, risks, public-good value, public-safe implications, safeguard implications, technical implications, data implications, AI implications, cyber implications, privacy implications, national implications, public authority learning implications, finance-readiness implications, procurement-neutrality implications, and lawful handoff implications of research outputs.

8.1.7.2 Research Impact Assessment shall consider not only positive impact claims but also misuse, overclaim, false certainty, harmful publication, protected knowledge exposure, sensitive geospatial exposure, cyber-sensitive disclosure, biosecurity-sensitive disclosure, community harm, Indigenous protocol breach where applicable, sponsor capture, provider capture, public authority boundary collapse, finance boundary overclaim, procurement boundary overclaim, and execution overclaim.

8.1.7.3 Research Impact Assessment shall be recorded and correctable. It shall not be treated as approval, certification, legal compliance determination, public authority decision, finance approval, insurance approval, procurement decision, consent, deployment authorization, or execution.

### 8.1.8 Research-to-Foundry Pathway.

8.1.8.1 Research-to-Foundry Pathway shall convert research outputs into Foundry-ready Docket items, Quests, Bounties, Builds, Micro-Production tasks, open technical baselines, software objects, data pipeline builds, dashboard builds, Studio workflow builds, Marketplace object builds, Registry record builds, Grid input builds, TRL evidence builds, Reports builds, Campaign builds, National Portfolio builds, Nexus Universe builds, and lawful handoff dependency builds.

8.1.8.2 A research item shall not enter Foundry unless it has a recorded purpose, scope, source class, evidence basis, data status, AI-use status, public-safe status, safeguard status, review needs, support class, release class expectation, correction pathway, and archive rule. Where research is uncertain, incomplete, sensitive, dual-use, protected, or public authority-sensitive, Foundry routing shall reflect those limits.

8.1.8.3 Research-to-Foundry transfer shall not create technical validation, certification, procurement status, financeability, insurance approval, public authority approval, consent, deployment authorization, or execution.

### 8.1.9 Research-to-Reports Pathway.

8.1.9.1 Research-to-Reports Pathway shall convert research into public-safe or controlled publications, including technical notes, research summaries, evidence briefs, data papers, software release papers, methods notes, policy learning notes, National Portfolio Reports, WFEH-B Reports, DRR Reports, DRF Reports, DRI Reports, DICE Reports, GRIx Reports, Observatory Reports, Foundry Reports, Campaign Reports, Academy Reports, Labs Reports, Nexus Universe Reports, Grid and TRL Reports, handoff context notes, correction reports, and archive reports.

8.1.9.2 Research-to-Reports routing shall include evidence basis, methodology, limitations, source status, data availability statement, AI-use statement, public-safe abstract, protected knowledge controls, sensitive geospatial controls, dual-use controls, sponsor and provider disclosures where applicable, correction pathway, withdrawal pathway, and archive rule.

8.1.9.3 Publication shall not create approval, public warning, certification, procurement recommendation, financeability, insurance approval, public authority decision, country ranking, consent, deployment authorization, or execution.

### 8.1.10 Research-to-Handoff Context Pathway.

8.1.10.1 Research-to-Handoff Context Pathway shall convert bounded research outputs into dependency context for competent lawful actors. Such context may include evidence basis, method basis, data basis, model basis, software basis, uncertainty, limitations, assumptions, dependencies, public authority dependencies, legal dependencies, safeguard dependencies, finance and insurance questions, procurement boundaries, provider-neutrality notes, sponsor-boundary notes, support status, correction pathway, recall pathway, and archive rule.

8.1.10.2 Research-to-Handoff Context shall be prepared only after appropriate review, classification, public-safe analysis, safeguard review, data and AI review where applicable, and handoff boundary review. It shall be clear that the research supplies context and dependencies only.

8.1.10.3 Research-to-Handoff Context shall not authorize implementation, procurement, finance, insurance, certification, public authority action, consent, deployment, operation, command, or execution.

## 8.2 Research Stream Classes

### 8.2.1 WFEH-B Research Streams.

8.2.1.1 WFEH-B Research Streams shall address water, food, energy, health, biodiversity, nature, and their cross-system dependencies, including resource security, resilience, infrastructure interdependence, climate stress, supply chains, cyber-physical risk, community vulnerability, public health resilience, biodiversity loss, ecosystem services, cascading failure, adaptive capacity, observability needs, DRI indicators, Studio scenarios, National Portfolio needs, and lawful handoff dependencies.

8.2.1.2 WFEH-B Streams shall preserve public-safe reporting, protected knowledge, sensitive geospatial controls, health data controls, community safeguards, Indigenous protocols where applicable, public authority boundaries, finance-readiness boundaries, insurance-readiness boundaries, procurement neutrality, and correctionability.

8.2.1.3 WFEH-B research outputs shall not create environmental certification, public health decision, resource allocation, public warning, public authority approval, financeability, insurance approval, procurement status, consent, deployment authorization, or execution.

### 8.2.2 DRR Research Streams.

8.2.2.1 DRR Research Streams shall address disaster risk reduction, multi-hazard risk, exposure, vulnerability, resilience capacity, prevention, mitigation, preparedness, crisis learning, recovery learning, reconstruction learning, community resilience, early-warning interfaces, public-safe communication, national risk governance learning, public authority learning, and National Portfolio priorities.

8.2.2.2 DRR Streams shall support DRI, GRIx, Nexus Observatory, Nexus Studio, Nexus Reports, Nexus Academy, Risk Academy, Nexus Campaigns, Nexus Universe, National Working Groups, Competence Cells, and lawful handoff context.

8.2.2.3 DRR research shall not create public warnings, emergency commands, official disaster plans, public authority decisions, procurement decisions, financeability, insurance approval, consent, deployment authorization, or execution.

### 8.2.3 DRF Research Streams.

8.2.3.1 DRF Research Streams shall address disaster-risk-finance learning, protection gaps, risk layering, insurance-readiness questions, public finance relevance questions, donor-readiness questions, capital-readability, resilience finance literacy, assumptions registers, dependency registers, diligence-gap registers, and no-reliance readiness-room materials.

8.2.3.2 DRF Streams shall preserve regulated-perimeter discipline. Research may make questions legible to capital readers, insurers, donors, and public finance readers, but shall not arrange finance, recommend investments, underwrite risk, issue ratings, solicit transactions, allocate public finance, or produce financial advice.

8.2.3.3 DRF research outputs shall not create financeability, bankability, insurability, underwriting, donor commitment, public finance allocation, investment advice, procurement status, public authority approval, consent, deployment authorization, or execution.

### 8.2.4 DRI Research Streams.

8.2.4.1 DRI Research Streams shall address disaster-risk-intelligence indicators, signals, dashboards, confidence labels, uncertainty labels, hotspot records, multi-hazard records, cascade records, risk interpretation, public-safe intelligence summaries, observability integration, Studio scenario integration, National DRI contributions, correction, and archive.

8.2.4.2 DRI Streams shall maintain no-warning, no-rating, no-insurance-score, no-investment-signal, no-public-authority-decision, and no-emergency-command boundaries. DRI research shall support public-good intelligence and learning, not operational warning authority.

8.2.4.3 DRI research outputs shall not be public warnings, insurance scores, investment signals, official forecasts, emergency commands, public authority decisions, procurement signals, deployment authorizations, or execution instructions.

### 8.2.5 AI and Agentic Systems Streams.

8.2.5.1 AI and Agentic Systems Streams shall address AI models, frontier models, retrieval systems, classifiers, forecasting systems, optimization systems, decision-support systems, agentic workflows, AI safety, human oversight, red-team methods, evaluation harnesses, benchmark cards, model cards, system cards, agent workflow cards, AI-use labels, data provenance, prompt-injection controls, tool-permission controls, bias and harm review, public-safe AI output, and AI incident response.

8.2.5.2 AI Streams shall preserve human review and shall prohibit automated high-stakes decisions by default within NAF. Agentic workflows shall be governed by no-command rules, no-write-back rules, tool permission controls, escalation triggers, kill-switch conditions, output review, and correction records.

8.2.5.3 AI research outputs shall not create AI safety certification, automated decision authority, public authority decision, professional qualification, procurement readiness, financeability, insurance approval, deployment authorization, or execution.

### 8.2.6 Cyber and Cyber-Physical Streams.

8.2.6.1 Cyber and Cyber-Physical Streams shall address cybersecurity, zero trust, identity and access, secure repositories, SBOMs, dependency risk, vulnerability disclosure, incident response, OT, IoT, IIoT, critical infrastructure, cyber-physical resilience, secure collaboration, cyber-sensitive publication, public-safe disclosure, and degraded-mode continuity.

8.2.6.2 Cyber Streams shall distinguish research, public-safe summaries, secure-room materials, vulnerability-sensitive materials, operationally sensitive materials, and handoff-recipient-only materials. Cyber-sensitive research shall be controlled to avoid enabling harm.

8.2.6.3 Cyber research outputs shall not create security certification, operational command, public warning, legal compliance determination, public authority approval, deployment authorization, or execution.

### 8.2.7 Geospatial, Earth Observation, and Digital Twin Streams.

8.2.7.1 Geospatial, Earth Observation, and Digital Twin Streams shall address geospatial layers, Earth observation records, remote sensing, spatial analytics, protected location controls, sensitive infrastructure masking, protected species masking, sacred site controls, community-sensitive location generalization, digital twin assumptions, scenario twins, infrastructure twins, WFEH-B twins, disaster-risk twins, sensor-linked twins, Studio workflows, public-safe atlas outputs, and handoff dependency records.

8.2.7.2 Such Streams shall maintain sensitive location controls, delay rules, aggregation rules, public-safe map publishing, protected knowledge controls, Indigenous protocol controls where applicable, cyber-sensitive controls, and public authority boundary controls.

8.2.7.3 Geospatial, Earth observation, and digital twin research outputs shall not create official maps, public warnings, surveillance authority, operational command, land-use approval, deployment authorization, or execution.

### 8.2.8 Telecom, AI-RAN, O-RAN, Edge, and Private Wireless Streams.

8.2.8.1 Telecom, AI-RAN, O-RAN, Edge, and Private Wireless Streams shall address network architecture records, interoperability profiles, spectrum dependency records, telecom regulatory dependencies, edge node records, AI-RAN methods, O-RAN interfaces, private wireless patterns, cyber-resilience controls, degraded-mode network operations, public authority dependencies, provider-neutrality records, Studio demonstrations, Nexus Universe Core Build needs, and handoff context.

8.2.8.2 Such Streams shall be standards-aware and provider-neutral. Research may map interfaces and dependencies, but shall not create standards authority, spectrum authority, telecom regulatory approval, provider validation, procurement recommendation, deployment authorization, or execution.

8.2.8.3 Telecom research outputs shall not authorize network deployment, spectrum use, public authority action, procurement, finance, insurance, certification, operational command, or execution.

### 8.2.9 Sovereign Compute, HPC, Cloud, and Quantum-Relevant Security Streams.

8.2.9.1 Sovereign Compute, HPC, Cloud, and Quantum-Relevant Security Streams shall address compute environment records, sovereign-sensitive workload classification, secure enclaves, confidential computing where appropriate, compute-to-data workflows, cloud and HPC architecture, verifiable compute, cost and support records, access controls, post-quantum transition literacy, cryptographic dependencies, national hosting context, cross-border transfer controls, continuity, teardown, and archive.

8.2.9.2 Such Streams shall preserve provider neutrality, data sovereignty, public authority-sensitive controls, national infrastructure boundaries, restricted-data controls, AI-use controls, cyber controls, and support-without-warranty discipline.

8.2.9.3 Compute research outputs shall not create provider validation, service warranty, public authority approval, data right, deployment authorization, financeability, procurement status, or execution.

### 8.2.10 Robotics, Drones, Sensors, IoT, OT, and IIoT Streams.

8.2.10.1 Robotics, Drones, Sensors, IoT, OT, and IIoT Streams shall address physical systems, sensing networks, edge devices, robotics constraints, drone constraints, telemetry, sensor records, industrial systems, cyber-physical interfaces, operational risk, safety dependencies, public authority dependencies, geospatial sensitivity, privacy, public-safe publication, Studio demonstrations, and handoff context.

8.2.10.2 Such Streams shall maintain safety, cyber, privacy, protected location, public authority, community, Indigenous protocol where applicable, and deployment boundary controls. Research shall remain distinguishable from operational deployment.

8.2.10.3 Research outputs in these Streams shall not create airspace approval, device certification, operational command, infrastructure access, surveillance authority, public authority approval, deployment authorization, or execution.

### 8.2.11 Biosecurity-Sensitive Governance Streams.

8.2.11.1 Biosecurity-Sensitive Governance Streams shall address governance, data, methods, public-safe publication, risk classification, dual-use concerns, protected knowledge, health-sensitive data, One Health learning, biosurveillance boundary discipline, public authority learning, public health boundary records, safeguard controls, and correction pathways.

8.2.11.2 Such Streams shall apply heightened review for dual-use sensitivity, harmful disclosure, sensitive protocols, protected knowledge, public health overclaim, public authority overclaim, and public-safe publication. Where necessary, outputs shall be restricted, summarized, delayed, routed to secure rooms, or withheld.

8.2.11.3 Biosecurity-sensitive research outputs shall not create medical advice, public health decision, regulatory approval, biosafety certification, public warning, operational instruction, deployment authorization, or execution.

### 8.2.12 Climate, Nature, Biodiversity, and Earth Systems Streams.

8.2.12.1 Climate, Nature, Biodiversity, and Earth Systems Streams shall address climate adaptation, climate risk, nature-based solutions, biodiversity monitoring, ecosystem services, Earth systems, protected species data, environmental monitoring, land and water systems, resilient infrastructure, community resilience, sensitive location controls, WFEH-B cascades, DRI indicators, Observatory signals, Studio scenarios, Reports, National Portfolios, and handoff dependencies.

8.2.12.2 Such Streams shall guard against greenwashing, false certainty, sensitive biodiversity data exposure, protected knowledge exposure, community harm, Indigenous protocol breaches where applicable, and public authority overclaim.

8.2.12.3 Climate, nature, biodiversity, and Earth systems research outputs shall not create environmental certification, regulatory approval, official forecasts, public authority decision, procurement preference, financeability, insurance approval, consent, deployment authorization, or execution.

### 8.2.13 Advanced Manufacturing, Semiconductors, and Supply-Chain Streams.

8.2.13.1 Advanced Manufacturing, Semiconductors, and Supply-Chain Streams shall address manufacturing readiness, semiconductor supply-chain context, critical materials, logistics, cyber-physical manufacturing systems, industrial resilience, export-control sensitivity, dual-use sensitivity, standards-interface records, quality methods, national capability records, workforce implications, public-good software and data objects, and handoff dependency records.

8.2.13.2 Such Streams shall preserve national security sensitivity, export-control awareness, IP boundaries, provider neutrality, procurement neutrality, sponsor boundaries, public authority dependencies, and lawful handoff discipline.

8.2.13.3 Research outputs in these Streams shall not create export authorization, product certification, supplier approval, procurement recommendation, financeability, insurance approval, public authority approval, deployment authorization, or execution.

## 8.3 Research Object Taxonomy

### 8.3.1 Research Questions.

8.3.1.1 Research Questions are recorded inquiries that define what must be investigated before a NAF item can progress. They may arise from signals, Dockets, National Portfolios, Working Groups, Competence Cells, Campaigns, Academy pathways, Foundry items, Studio workflows, Reports, DRI indicators, Observatory signals, Grid reviews, TRL reviews, Nexus Universe outputs, standards-interface gaps, or handoff dependency needs.

8.3.1.2 A Research Question shall include purpose, scope, source, relevance, domain, assumptions, constraints, expected output, review need, public-safe status, data status, AI-use status, safeguard status, standards-interface relevance where applicable, national relevance where applicable, correction pathway, and archive rule.

8.3.1.3 A Research Question is not evidence, approval, certification, public authority decision, finance signal, procurement signal, consent, deployment authorization, or execution.

### 8.3.2 Evidence Gaps.

8.3.2.1 Evidence Gaps are recorded deficiencies, uncertainties, missing data, missing methods, weak assumptions, unresolved dependencies, unreviewed claims, insufficient benchmarks, unvalidated models, incomplete public-safe summaries, missing safeguards, missing standards-interface records, missing national context, or missing handoff dependencies.

8.3.2.2 Evidence Gaps shall be routed to Research Quests, Evidence Quests, Data Quests, Labs Streams, Foundry tasks, Reports review, Studio review, Grid review, TRL note preparation, National Portfolio review, Nexus Universe preparation, or correction.

8.3.2.3 Identification of an Evidence Gap shall not be treated as failure by itself; it is a validity-by-record function. Unrecorded gaps are more dangerous than recorded uncertainty.

### 8.3.3 Testbed Records.

8.3.3.1 Testbed Records shall document controlled environments, experimental setups, simulation environments, data environments, software environments, compute environments, telecom environments, sensor environments, cyber environments, digital twin environments, Studio environments, or field-adjacent environments used for research and evaluation.

8.3.3.2 Testbed Records shall include purpose, scope, environment, configuration, data status, AI-use status, access controls, safety controls, cyber controls, public-safe status, safeguard status, limitations, repeatability status, review status, correction pathway, teardown rule, and archive rule.

8.3.3.3 Testbed Records shall not create deployment authorization, operational approval, public authority approval, safety certification, financeability, procurement readiness, or execution.

### 8.3.4 Method Notes.

8.3.4.1 Method Notes shall document research methods, evidence methods, data methods, AI methods, cyber methods, observability methods, DRI methods, GRIx methods, Studio methods, evaluation methods, benchmark methods, public-safe transformation methods, safeguard methods, and handoff dependency methods.

8.3.4.2 Method Notes shall include purpose, scope, procedure, assumptions, limitations, required inputs, prohibited uses, data and AI labels, reproducibility status, review status, public-safe status, safeguard status, version, correction pathway, and archive rule.

8.3.4.3 A Method Note is not certification, standards authority, legal approval, public authority decision, procurement approval, deployment authorization, or execution.

### 8.3.5 Dataset Records.

8.3.5.1 Dataset Records shall document datasets used or produced in R\&D, including raw data, processed data, public-safe data, aggregated data, synthetic data, metadata-only records, data dictionaries, codebooks, schemas, feature sets, benchmark datasets, DRI datasets, Observatory datasets, Studio datasets, National Portfolio datasets, and handoff-recipient-only datasets.

8.3.5.2 Dataset Records shall include provenance, rights, license, sensitivity, data-use label, AI-use label, lineage, quality, limitations, public-safe status, access class, retention, deletion, sealing, correction pathway, and archive rule.

8.3.5.3 Dataset Records shall not create unrestricted data rights, AI training permission, public authority approval, handoff permission, consent, deployment authorization, or execution.

### 8.3.6 Model Cards.

8.3.6.1 Model Cards shall document AI, statistical, simulation, forecasting, classification, optimization, risk, digital twin, benchmark, and evaluation models used or produced in R\&D. Model Cards shall include model purpose, intended use, prohibited use, data provenance, training or calibration data constraints, evaluation records, bias and harm review, limitations, uncertainty, human review requirements, AI-use label, incident pathway, correction pathway, withdrawal rule, and archive rule.

8.3.6.2 Model Cards shall be required where R\&D outputs rely on models in a way that affects evidence, public-safe reporting, Studio workflows, DRI indicators, Observatory interpretation, Grid inputs, TRL notes, National Portfolios, or handoff context.

8.3.6.3 A Model Card is not AI safety certification, deployment approval, public authority decision, financeability, insurance approval, procurement readiness, or execution.

### 8.3.7 System Cards.

8.3.7.1 System Cards shall document systems composed of models, data, software, APIs, workflows, dashboards, infrastructure, human review, governance controls, access controls, and output review. System Cards shall be used for AI systems, Studio systems, Observatory systems, DRI systems, data systems, compute systems, dashboard systems, and handoff-relevant systems.

8.3.7.2 System Cards shall include system purpose, architecture, components, dependencies, data flows, AI-use status, cyber controls, privacy controls, human review, public-safe controls, safeguard controls, limitations, support class, release class, incident pathway, correction pathway, and archive rule.

8.3.7.3 A System Card is not system certification, security certification, public authority approval, procurement approval, deployment authorization, warranty, or execution.

### 8.3.8 Benchmark Records.

8.3.8.1 Benchmark Records shall document benchmark tasks, datasets, metrics, evaluation methods, performance results, limitations, test conditions, reproducibility status, uncertainty, bias and harm considerations, public-safe interpretation, and correction pathway.

8.3.8.2 Benchmark Records may support model review, software review, data review, Grid inputs, TRL notes, Studio workflows, Reports, Marketplace objects, Registry records, and handoff dependency context.

8.3.8.3 Benchmark Records shall not be treated as general validation, certification, procurement readiness, financeability, insurance approval, public authority approval, deployment authorization, or execution.

### 8.3.9 Scenario Records.

8.3.9.1 Scenario Records shall document scenarios used for learning, simulation, digital twins, stress testing, DRI interpretation, public authority learning, readiness rooms, finance-readiness questions, insurance-readiness questions, National Portfolio review, Nexus Universe arenas, and handoff dependency review.

8.3.9.2 Scenario Records shall include purpose, assumptions, inputs, scenario class, limitations, uncertainty, confidence where applicable, data status, AI-use status, public-safe status, safeguard status, interpretation rules, prohibited uses, correction pathway, and archive rule.

8.3.9.3 Scenario Records are not forecasts, public warnings, official plans, public authority decisions, finance decisions, insurance decisions, procurement decisions, deployment authorizations, or execution instructions.

### 8.3.10 Public-Safe Summaries.

8.3.10.1 Public-Safe Summaries shall translate research into accessible, bounded, non-overclaiming, non-sensitive, non-executing public language. They shall be used for Reports, Campaigns, Marketplace listings, Registry public fields, Academy materials, Nexus Universe materials, National Portfolio summaries, DRI summaries, Observatory summaries, and public-safe handoff context.

8.3.10.2 Public-Safe Summaries shall remove or control public warning overclaim, approval overclaim, finance overclaim, procurement overclaim, certification overclaim, consent overclaim, execution overclaim, protected knowledge, sensitive geospatial details, cyber-sensitive details, biosecurity-sensitive details, and other harmful disclosure.

8.3.10.3 A Public-Safe Summary is not public warning, approval, endorsement, certification, financeability, procurement status, consent, deployment authorization, or execution.

### 8.3.11 Research Impact Records.

8.3.11.1 Research Impact Records shall document expected and potential effects of research, including public-good value, learning value, evidence value, risk-reduction value, resilience value, national capability value, technology value, public-safe value, safeguard implications, misuse risks, dual-use risks, data risks, AI risks, cyber risks, public authority boundary risks, finance/procurement boundary risks, and handoff implications.

8.3.11.2 Research Impact Records shall support portfolio prioritization, public-safe review, safeguard review, Foundry routing, Reports routing, Studio routing, Grid and TRL routing, Nexus Universe preparation, National Portfolio memory, and handoff dependency review.

8.3.11.3 Research Impact Records shall not be impact certification, public authority approval, financeability, procurement readiness, endorsement, consent, deployment authorization, or execution.

### 8.3.12 Foundry Transfer Records.

8.3.12.1 Foundry Transfer Records shall document the movement of research into Foundry work. They shall identify the research source, Docket item, build opportunity, Quest or Bounty structure, expected output, evidence basis, unresolved gaps, data status, AI-use status, cyber status, public-safe status, safeguard status, support class, release class expectation, maintainer needs, review gates, correction pathway, and archive rule.

8.3.12.2 Foundry Transfer Records shall make explicit what is ready for build and what remains unresolved. They shall prevent research uncertainty from being misread as readiness.

8.3.12.3 Foundry Transfer Records shall not create validation, certification, procurement status, financeability, insurance approval, public authority approval, consent, deployment authorization, or execution.

## 8.4 R\&D Governance

### 8.4.1 Research Ethics Boundaries.

8.4.1.1 R\&D under NAF shall observe research ethics boundaries appropriate to the subject matter, participants, data, communities, jurisdictions, institutions, technologies, and risks involved. Ethics boundaries may include human participant protections, privacy protections, youth protections, community safeguards, Indigenous protocols where applicable, protected knowledge controls, dual-use review, public-safe publication, conflict disclosure, and correction rights.

8.4.1.2 Where institutional ethics approval, public authority permission, community process, Indigenous process, data permission, biosafety review, cybersecurity control, export-control review, or other external lawful process is required, NAF research shall not proceed beyond permissible preparatory work without appropriate recorded authority.

8.4.1.3 NAF ethics review or boundary review shall not replace legally required ethics approval, public authority approval, community consent, Indigenous consent, institutional review, regulatory approval, or professional duty.

### 8.4.2 Data Rights.

8.4.2.1 R\&D shall respect data rights, including source rights, license restrictions, privacy rights, data subject rights where applicable, community data rights, Indigenous data sovereignty where applicable, public authority data restrictions, health data restrictions, youth data restrictions, cyber-sensitive restrictions, infrastructure-sensitive restrictions, protected knowledge restrictions, geospatial restrictions, and cross-border transfer rules.

8.4.2.2 Data used in R\&D shall carry data-use labels, AI-use labels, access classes, sensitivity classes, provenance records, rights records, retention rules, deletion rules, sealing rules, correction pathways, and archive rules.

8.4.2.3 Access to data for R\&D shall not create ownership, unrestricted reuse, publication permission, AI training permission, handoff permission, public authority approval, consent, deployment authorization, or execution.

### 8.4.3 AI-Use Controls.

8.4.3.1 AI use in R\&D shall be governed by recorded AI-use labels, data-use labels, human review, model cards, system cards, benchmark records, agent workflow cards where applicable, prohibited use rules, prompt-injection controls, tool-permission controls, data leakage controls, bias and harm review, public-safe output review, incident response, correction, withdrawal, and archive.

8.4.3.2 R\&D shall not use AI to make high-stakes decisions by default. AI may support retrieval, summarization, classification, evaluation, drafting, coding, analysis, simulation, benchmarking, or workflow support only within recorded controls and human review requirements.

8.4.3.3 AI-enabled R\&D outputs shall not create automated authority, AI safety certification, public authority decision, professional determination, financeability, insurance approval, procurement status, deployment authorization, or execution.

### 8.4.4 Public-Safe Publication.

8.4.4.1 R\&D publication shall be governed by public-safe publication controls. These controls shall determine whether research may be open public, controlled public, public-safe summary, repository deposit, metadata-only record, secure-room-only, data-room-only, handoff-recipient-only, delayed, withdrawn, or archived.

8.4.4.2 Public-Safe Publication shall control public warning language, public authority overclaim, finance overclaim, procurement overclaim, certification overclaim, consent overclaim, execution overclaim, protected knowledge exposure, sensitive geospatial exposure, cyber-sensitive exposure, biosecurity-sensitive exposure, and harmful technical detail.

8.4.4.3 Publication shall not create approval, endorsement, public warning, certification, procurement recommendation, financeability, insurance approval, public authority decision, consent, deployment authorization, or execution.

### 8.4.5 Protected Knowledge Controls.

8.4.5.1 Protected Knowledge Controls shall apply to knowledge that is culturally sensitive, community-sensitive, Indigenous protocol-sensitive where applicable, security-sensitive, cyber-sensitive, biosecurity-sensitive, health-sensitive, infrastructure-sensitive, geospatial-sensitive, commercially confidential, legally restricted, or otherwise unsuitable for open publication.

8.4.5.2 Protected Knowledge shall be classified, access-controlled, transformed, summarized, delayed, restricted, sealed, or excluded as appropriate. Public-safe summaries may be produced only where they do not expose protected content or enable harm.

8.4.5.3 Protected Knowledge controls shall not imply ownership transfer, consent, permission to publish, permission to train AI, public authority approval, deployment authorization, or execution.

### 8.4.6 Community Safeguards.

8.4.6.1 R\&D involving communities shall be non-extractive, respectful, recorded, public-safe, rights-aware, accessibility-aware, language-aware, and correctionable. Community participation shall be documented with scope, purpose, data conditions, protected knowledge conditions, consent boundary records, public-safe status, and correction pathway.

8.4.6.2 Community safeguards shall apply to community knowledge, lived-risk knowledge, local context, vulnerability information, resilience priorities, public-safe reporting, National Portfolio inputs, Campaign inputs, Observatory needs, DRI interpretation, Studio participation, and handoff context.

8.4.6.3 Community participation in R\&D shall not imply community consent, endorsement, land access, project approval, deployment approval, data rights beyond scope, or execution authority.

### 8.4.7 Public Authority Boundaries.

8.4.7.1 R\&D may support public authority learning, policy intelligence, evidence review, standards-interface dialogue, disaster-risk learning, public finance learning, regulatory learning, and National Portfolio context, but shall not substitute for public authority decisions.

8.4.7.2 Public authority participation in R\&D shall be recorded as learning or review unless separately and lawfully recorded otherwise by the competent public authority. Public authority materials shall be classified for sensitivity, confidentiality, public-safe status, data status, and archive status.

8.4.7.3 R\&D shall not create official action, public policy adoption, regulation, permit, license, public finance allocation, procurement decision, public warning, emergency command, deployment authorization, or execution.

### 8.4.8 Sponsor and Provider Controls.

8.4.8.1 Sponsor and Provider participation in R\&D shall be governed by support-without-control, provider contribution without validation, no-pay-to-influence, no-pay-to-routing, no-pay-to-validation, public-good firewall, conflict disclosure, transparency, IP and license clarity, public-safe claims, correction, and archive.

8.4.8.2 Sponsor support or provider contribution may support research resources, infrastructure, compute, connectivity, software, data, expertise, venues, learning pathways, testbeds, or Nexus Universe demonstrations, but shall not control research questions, methods, findings, review outcomes, public-safe release decisions, Registry status, Marketplace listing, Grid inputs, TRL notes, or handoff conditions except through recorded and bounded contribution roles.

8.4.8.3 Sponsor or provider involvement shall not imply endorsement, validation, certification, procurement preference, financeability, insurance approval, public authority approval, deployment authorization, or execution.

### 8.4.9 Dual-Use Controls.

8.4.9.1 Dual-Use Controls shall apply where R\&D outputs may be used for harmful purposes, including cyber misuse, biosecurity misuse, surveillance misuse, infrastructure attack, weapons-related use, evasion, harmful automation, sensitive geospatial misuse, protected knowledge misuse, or other foreseeable harm.

8.4.9.2 Dual-use review may require restriction, redaction, delay, secure-room routing, public-safe transformation, access limitation, additional review, legal escalation, public authority-sensitive handling, withdrawal, recall, or archive.

8.4.9.3 Dual-use review shall not itself create public authority approval, security certification, publication permission, legal compliance determination, deployment authorization, or execution authority. It is a harm-prevention and boundary-preservation control.

### 8.4.10 Correction and Archive.

8.4.10.1 R\&D outputs shall be correctable throughout their lifecycle. Correction may include erratum, addendum, revision, supersession, downgrade, suspension, withdrawal, retraction where necessary, recall, public repair, access restriction, public-safe notice, Registry update, Marketplace update, Report correction, Studio restriction, Grid correction, TRL correction, handoff recall, archive, or non-continuation.

8.4.10.2 R\&D Archives shall preserve research questions, evidence gaps, methods, datasets, models, benchmarks, scenarios, public-safe summaries, impact records, transfer records, reviews, corrections, withdrawn materials, non-continuing items, and successor links. Archive records shall include archive-not-current notices where appropriate.

8.4.10.3 Correction and archive shall be treated as research integrity functions. A corrected research record is more trustworthy than an uncorrectable claim.

## 8.5 Research Translation

### 8.5.1 Science-to-Policy.

8.5.1.1 Science-to-Policy translation shall convert research into public authority learning materials, policy intelligence notes, legislative or regulatory learning context, standards-interface records, public finance learning questions, resilience planning inputs, National Portfolio context, and public-safe Reports without substituting for public authority decisions.

8.5.1.2 Science-to-Policy outputs shall identify evidence basis, uncertainty, limitations, jurisdictional limits, public authority dependencies, data status, AI-use status, public-safe status, safeguard status, implementation dependencies, correction pathway, and archive rule.

8.5.1.3 Science-to-Policy translation shall not create policy adoption, regulation, permit, license, public finance allocation, procurement decision, public warning, emergency command, public authority approval, deployment authorization, or execution.

### 8.5.2 Science-to-Community.

8.5.2.1 Science-to-Community translation shall convert research into accessible, respectful, non-extractive, public-safe, language-aware, culturally aware, disability-accessible, and community-correctable materials. It may support Campaigns, Academy pathways, community briefings, public-safe summaries, Studio learning, National Portfolio input, Observatory needs, DRI interpretation, and correction.

8.5.2.2 Science-to-Community outputs shall preserve community safeguards, Indigenous protocols where applicable, protected knowledge, consent boundaries, data boundaries, sensitive location controls, public-safe communication, and community-facing correction.

8.5.2.3 Science-to-Community translation shall not create community consent, Indigenous consent, endorsement, land access, project approval, deployment authorization, or execution.

### 8.5.3 Science-to-Technology.

8.5.3.1 Science-to-Technology translation shall convert research into technical requirements, methods, software builds, data pipelines, models, APIs, dashboards, digital twins, simulations, test suites, standards-interface mappings, open technical baselines, reference implementations, Studio workflows, Grid inputs, TRL notes, and handoff dependency context.

8.5.3.2 Science-to-Technology translation shall preserve evidence status, maturity limits, data rights, AI-use controls, cyber controls, privacy controls, safeguard controls, support class, release class, correction pathway, and archive rule.

8.5.3.3 Science-to-Technology translation shall not create technology certification, product approval, security approval, AI safety approval, procurement status, financeability, insurance approval, deployment authorization, or execution.

### 8.5.4 Science-to-Readiness.

8.5.4.1 Science-to-Readiness translation shall convert research into bounded readiness context, including evidence readiness, data readiness, technical readiness, public-safe readiness, safeguard readiness, national readiness, Universe readiness, finance-readiness questions, insurance-readiness questions, donor-readiness questions, public finance relevance questions, Grid inputs, TRL notes, and handoff dependency records.

8.5.4.2 Readiness translation shall record what is known, what is uncertain, what is unsupported, what requires review, what requires public authority action, what requires legal review, what requires finance or insurance diligence, what requires community process, what requires Indigenous process where applicable, and what requires correction.

8.5.4.3 Science-to-Readiness translation shall not create certification, maturity certification, procurement readiness, financeability, insurability, public authority approval, consent, deployment authorization, or execution.

### 8.5.5 Science-to-Public-Safe Reporting.

8.5.5.1 Science-to-Public-Safe Reporting translation shall convert research into public-safe summaries, Reports, briefings, Campaign materials, Academy materials, public dashboards, Marketplace descriptions, Registry public fields, Nexus Universe materials, and public-facing correction notices.

8.5.5.2 Such translation shall remove or control false certainty, public warning overclaim, approval overclaim, finance overclaim, procurement overclaim, certification overclaim, consent overclaim, execution overclaim, sensitive geospatial disclosure, cyber-sensitive disclosure, biosecurity-sensitive disclosure, protected knowledge disclosure, and harmful technical detail.

8.5.5.3 Public-safe reporting shall not be public warning, approval, endorsement, certification, financeability, procurement status, consent, deployment authorization, or execution.

### 8.5.6 Science-to-National Portfolios.

8.5.6.1 Science-to-National Portfolios translation shall convert research into National Context Records, National Systems-Risk Maps, National Challenge Briefs, Evidence Need Records, Observatory Need Records, Core Build Requests, Safeguard Records, Public Authority Learning Records, Readiness Question Records, Competence Cell Workplans, Universe Arena Routing Records, National Portfolio Reports, and National Portfolio Archives.

8.5.6.2 Such translation shall preserve national ownership, localization, language access, data sovereignty, public authority boundaries, community safeguards, Indigenous protocols where applicable, provider neutrality, sponsor boundaries, finance-readiness boundaries, procurement neutrality, public-safe reporting, correction, and archive.

8.5.6.3 Science-to-National Portfolios translation shall not create country ranking, government approval, public finance allocation, procurement plan, financeability, insurance approval, consent, deployment authorization, or execution.

### 8.5.7 Science-to-Lawful Handoff.

8.5.7.1 Science-to-Lawful Handoff translation shall convert research into dependency context for downstream lawful actors, including National Consortium Companies, Project SPVs, public authorities, providers, operators, contractors, funders, insurers, donors, universities, labs, communities where appropriate, and other competent lawful actors.

8.5.7.2 Such translation shall include evidence context, data context, method context, technical context, Studio context, Grid and TRL context, public-safe status, safeguard status, public authority dependencies, legal dependencies, finance and insurance questions, procurement boundaries, provider-neutrality notes, sponsor-boundary notes, recipient responsibilities, correction pathway, recall pathway, and archive rule.

8.5.7.3 Science-to-Lawful Handoff translation shall not authorize implementation, procurement, finance, insurance, certification, public authority action, consent, deployment, operation, command, or execution. It transfers dependencies, not authority.

## 8.6 R\&D Boundaries

### 8.6.1 Research Is Not Approval.

8.6.1.1 Research conducted, reviewed, published, presented, listed, recorded, cited, translated, or routed within NAF shall not be treated as approval of a technology, provider, sponsor, project, policy, method, dataset, model, dashboard, digital twin, Studio workflow, National Portfolio item, Nexus Universe output, Marketplace listing, Registry record, Grid input, TRL note, or handoff dependency package.

8.6.1.2 Approval, where required, must be granted separately by a competent public authority, institution, regulator, standards body, ethics body, procurement body, finance actor, insurer, community process, Indigenous process where applicable, enterprise actor, National Consortium Company, Project SPV, or other lawful actor with proper authority.

### 8.6.2 Testbed Is Not Deployment.

8.6.2.1 Testbeds, simulations, labs, sandboxes, pilots, demonstrations, controlled environments, secure rooms, data rooms, Studio workflows, digital twins, benchmarks, and Nexus Universe demonstrations shall not be treated as deployment unless separately and lawfully authorized by competent actors.

8.6.2.2 A Testbed may generate evidence, learning, methods, data, software, dashboards, scenarios, Grid inputs, TRL notes, Reports, and handoff context. It shall not create operational authority, production approval, public authority approval, safety certification, procurement status, financeability, insurance approval, consent, or execution.

### 8.6.3 Method Is Not Certification.

8.6.3.1 A research method, method note, ontology, schema, protocol mapping, benchmark method, evaluation method, public-safe method, data method, AI method, cyber method, Studio method, DRI method, GRIx method, Observatory method, Grid method, or TRL method shall not be treated as certification.

8.6.3.2 Methods may improve consistency, transparency, repeatability, reviewability, interoperability, public-safe reporting, and correctionability. They do not create certified status, compliance determination, standards conformance, procurement qualification, financeability, insurance approval, public authority approval, or deployment authorization.

### 8.6.4 Study Is Not Public Authority Decision.

8.6.4.1 A study, research report, policy note, evidence brief, DRI summary, Observatory summary, National Portfolio Report, Studio scenario, public authority learning record, readiness-room material, or Nexus Universe presentation shall not be treated as a public authority decision.

8.6.4.2 Public authority decisions remain outside NAF unless separately and lawfully made by the competent public authority through its own procedures. NAF may support learning and context; it shall not substitute for official action.

### 8.6.5 Research Output Is Not Financeability.

8.6.5.1 A research output, evidence pack, method note, model card, benchmark record, dataset record, Studio workflow, Grid input, TRL note, Report, Marketplace listing, Registry record, National Portfolio object, Nexus Universe output, or handoff dependency note shall not be treated as financeable, bankable, investible, insurable, underwritable, donor-approved, public-finance-approved, rated, or transaction-ready by implication.

8.6.5.2 Finance, insurance, donor, and public finance decisions require separate lawful processes, independent diligence, regulated-perimeter controls, no-reliance discipline, and competent actor decisions outside NAF’s default R\&D posture.

### 8.6.6 Research Participation Is Not Consent.

8.6.6.1 Participation in R\&D by communities, Indigenous participants where applicable, public authorities, universities, providers, sponsors, capital readers, insurers, donors, civil society, youth, students, workers, experts, or other stakeholders shall not be treated as consent, endorsement, approval, authorization, procurement support, finance support, insurance support, public authority action, deployment approval, or execution.

8.6.6.2 Consent, where required, must be separately and lawfully recorded through the appropriate process. Research participation records shall include consent boundary notices where participation could be misread as consent.

### 8.6.7 Lab Affiliation Is Not Endorsement.

8.6.7.1 Affiliation with Nexus Labs, Frontier Labs, university labs, public-sector labs, corporate labs, open-source labs, community labs, National Working Groups, Competence Cells, or Nexus Universe research environments shall not imply endorsement by Nexus, GCRI, The Global Risks Forum (GRF), The Global Risks Alliance (GRA), any public authority, any university, any sponsor, any provider, any capital reader, any insurer, any donor, any community, or any Indigenous institution.

8.6.7.2 Lab affiliation shall be recorded for role, scope, contribution, support, conflict, and correction purposes only. It shall not create certification, approval, procurement status, financeability, insurance approval, public authority approval, consent, deployment authorization, agency, employment, or execution.

8.6.7.3 The final R\&D rule of Part VIII is that NAF shall turn research into disciplined public-good records, methods, learning, builds, Reports, Studio workflows, Grid inputs, TRL notes, National Portfolio memory, Nexus Universe readiness, and lawful handoff context without converting research into authority. Research may discover, explain, test, model, simulate, translate, publish, teach, build, and correct; it shall not approve, certify, procure, finance, insure, warn, consent, deploy, command, operate, or execute by implication.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.therisk.global/organization/operations/frameworks/nexus-agile-framework-naf/viii.-r-and-d.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.