Included
- Controlling SCALE agent instructions
- Copy-and-paste launch prompt
- Vendor-neutral platform setup guide
- Analysis workbook and reference architecture
- Markdown, JSON, and CSV workspace formats
- Fictional source case and completed walkthrough
SCALE Agent Kit · Bring your own AI
Bring SCALE to the AI you already use.
Strategic Capabilities Architecture Linkage Engine is a downloadable Business Architecture agent kit. Load it into an AI environment approved by your organization, supply a bounded question and authorized evidence, then use SCALE to structure architecture, trace relationships, expose uncertainty, and create decision-focused work. The business architect directs the analysis and retains professional judgment.
Portable agent kit
One download turns your approved AI into a SCALE workbench.
The package supplies the operating instructions, launch prompt, workspace formats, analysis method, reference model, and worked example. You supply the AI environment, the business question, the authorized evidence, and the professional judgment.
Four-step setup
Download, load, launch, and work.
Use a persistent project or custom-agent workspace when your AI supports one. A bounded conversation with file attachments also works when you export the workspace before ending.
Get one portable kit
Download the ZIP and keep the files together. No SCALE account, hosted service, or vendor-specific installation is required.
Bring it to your AI
Place the agent instructions in the strongest instruction field your approved AI provides, then attach the reference and schema files.
Paste the starting command
Use the supplied launch prompt to establish the question, decision owner, evidence boundary, work product, and stop conditions.
Review and export
Direct the analysis, disposition proposals, create the work product, and export the Markdown and JSON workspace for governed continuation.
Security boundary: use only an AI environment approved for the content you intend to provide. SCALE can record access classifications, but a portable instruction kit cannot enforce the permissions, retention, training-use, or records controls of the selected platform.
Inside the package
Everything needed to start without inventing the agent.
The kit separates controlling instructions, architecture method, portable data structures, and teaching material so evidence cannot silently redefine how the agent operates.
Agent instructions
The controlling role, authority, workflow, review rules, commands, and answer contract.
Launch and setup guides
A copy-and-paste launch prompt plus vendor-neutral setup patterns for projects, custom agents, or one-time conversations.
Portable workspace
A human-readable export template, JSON schema, starter JSON, and CSV registers for evidence, architecture, analysis, proposals, and decisions.
Methods and example
The analysis workbook, version 0.5 reference architecture, and a completed Northstar walkthrough with source case.
Product proposition
A portable kit, not a hosted or automated architect.
Business architecture work rarely begins with a document. It begins with a strategy, initiative, decision, problem, or change that somebody needs to understand. The relevant knowledge is usually distributed across models, repositories, portfolio systems, policies, requirements, documents, decisions, and experienced people.
SCALE supplies a disciplined agent operating model and portable workspace files for connecting that evidence, structuring the relevant architecture, investigating relationships, and producing decision support inside the user’s chosen AI. Exports preserve accepted concepts, evidence paths, corrections, and institutional decisions outside conversational memory.
AI is the mechanism, not the authority or the proposition. SCALE can search, synthesize, propose, compare, trace, challenge, and summarize. It cannot determine enterprise truth, approve architecture, silently reconcile conflicts, or take decision rights from accountable people.
1 ZIPThe complete portable kit
Your AIThe execution environment
BAPrimary user and operator
MD + JSONPortable workspace exports
First use
What are you trying to understand?
Choose the business-architecture purpose before selecting sources or tools. The question establishes which architecture and evidence are relevant.
Assess an initiative
Understand affected capabilities, value streams, organizations, information, stakeholders, dependencies, and outcomes.
Explore a strategy
Trace strategic intent into the parts of the enterprise required to execute it and expose unsupported assumptions.
Analyze a change
Identify business impacts, gaps, dependencies, constraints, and the architecture questions that still require investigation.
Build architecture
Develop or extend capability, value-stream, organization, information, stakeholder, strategy, and initiative views.
Create a work product
Use existing architecture and enterprise evidence to prepare a decision-focused assessment, brief, heatmap, or cross-map.
Design rule: SCALE begins with a business architect defining the decision context and the architecture needed to support it. Sources and tools follow from that purpose.
Decision-led path
Question → Evidence → Architecture → Analysis → Work product
The sequence creates immediate decision value while leaving behind governed architecture that can support the next question.
Establish the decision context
The business architect defines what must be understood, for whom, by when, within what scope, and what a sufficient answer must contain.
Assemble governed evidence
Connect the documents, source records, existing models, decisions, and experienced owners relevant to the question.
Develop the relevant structure
Identify and relate capabilities, value streams, organizations, information, stakeholders, strategies, initiatives, and other necessary concepts.
Trace implications and uncertainty
Evaluate impacts, gaps, dependencies, overlap, conflicts, confidence, and unresolved architecture questions.
Prepare usable decision support
Turn the governed result into the view, assessment, options, and next decision required by the audience.
Continuous architecture work
Architecture accumulates through five recurring activities.
SCALE supports an ongoing BA operating model. Each analysis can extend the workspace, improve its evidence, and make later work faster and more coherent.
Bring evidence into context
Find strategies, initiatives, processes, organizational information, existing architecture, requirements, decisions, and other enterprise evidence relevant to the question.
Organize architecture meaning
Propose concepts and typed relationships for business-architect review without presenting machine interpretation as approved enterprise truth.
Investigate the architecture
Trace impacts, gaps, duplication, dependencies, alignment, capability implications, affected stakeholders, and questions worth pursuing.
Produce usable BA work
Develop impact assessments, initiative decompositions, capability assessments, alignment views, heatmaps, briefs, and other decision-focused work products.
Keep the architecture alive
Refine mappings, add evidence, challenge conclusions, resolve ownership, record decisions, and carry accepted knowledge into later work.
The continuity principle: analyses conclude, but validated architecture, evidence, review history, and unresolved questions remain available for governed reuse.
Human and agent responsibilities
The division of labor is explicit.
SCALE accelerates the work without obscuring who interprets the enterprise, who validates its meaning, and who has authority to decide.
Three-layer product model
Persistent knowledge, governed assistance, practitioner work.
In the portable kit, the workspace is the exported state, the agent is the controlling instruction set, and the workbench is the approved AI project or conversation where the architect performs the work.
Layer 01 · Persistent knowledge
Architecture Workspace
The portable Markdown, JSON, and CSV state that carries accepted architecture, evidence, decisions, and open items across AI sessions.
- Capabilities and value streams
- Organizations and stakeholders
- Information and business objects
- Strategies and initiatives
- Typed relationships, evidence, status, ownership, and history
Layer 02 · Reasoning assistance
SCALE Agent
The controlling instruction file that directs the user’s approved AI to work over supplied evidence while keeping uncertainty and source authority visible.
- Extract and suggest
- Compare and trace
- Challenge and analyze
- Detect inconsistencies
- Summarize with provenance
Layer 03 · Practitioner work
Business Architecture Workbench
The project or conversation in the user’s AI where the business architect performs analysis, applies judgment, and composes decision support.
- Impact and initiative analysis
- Strategy and capability alignment
- Modernization and operating-model analysis
- Stakeholder and information analysis
- Architecture briefs and work-product creation
Business Architecture Workbench
Organized around BA work, not model capabilities.
The conversation stays organized around recognizable practitioner analyses and work products. Extract, compare, and summarize remain supporting actions inside those tasks.
Impact analysis
Trace a proposed change across value, capabilities, information, organization, stakeholders, initiatives, policies, and enablement.
Initiative analysis
Clarify intended outcomes, affected architecture, dependencies, duplication, gaps, sequencing, and ownership.
Strategy alignment
Connect strategic intent to required value delivery, capability change, measures, investments, and accountable owners.
Capability assessment
Combine performance, maturity, strategic importance, pain points, evidence, and investment coverage without reducing the analysis to a score.
Modernization analysis
Translate technology-centered change into business-domain, value-stream, capability, information, organization, and stakeholder implications.
Operating-model analysis
Examine how capabilities, structure, decision rights, information, governance, and enabling resources work together.
Stakeholder analysis
Identify who receives value, performs work, owns meaning, controls policy, bears risk, and has formal decision rights.
Architecture brief
Compose a bounded decision view with findings, evidence, confidence, implications, options, owners, and the next governed action.
Architecture with provenance
A finding is incomplete until the architect can inspect its basis.
SCALE should make a claim, its evidence path, its uncertainty, and its human disposition inseparable. Provenance is not a citation added after the answer. It is part of the architecture.
Direct finding
A concise response to the business question, bounded by the requested scope and time horizon.
Architecture path
The typed relationship chain followed, such as objective → outcome → value stage → capability → initiative → owner.
Evidence
The governed records and source passages supporting every material enterprise claim.
Conflicts and gaps
Contradictory definitions, stale records, missing owners, broken links, and evidence the workspace does not contain.
Confidence
A visible assessment based on source authority, freshness, agreement, completeness, and retrieval quality.
Decision and owner
The judgment still required, the actor or forum authorized to make it, and the next governed action.
Governance model
Propose, disclose, review, and record.
The operating model preserves a clear line between assistance, professional interpretation, content stewardship, and formal decision authority.
Evidence before fluency
A polished response without traceable enterprise evidence is a liability. Material claims require governed support or an explicit non-fact label.
Proposals are not approvals
SCALE can identify a candidate concept, relationship, correction, or conclusion. The business architect and accountable stewards determine its disposition.
Competing truths stay visible
Conflicting definitions, local variation, exceptions, and unresolved ownership are governed conditions to expose, not noise to average away.
Human authority remains
The business architect owns the analysis and resulting architecture; accountable business owners and forums retain formal decision rights.
Authorization remains external
The user must bring only evidence authorized for the chosen AI environment. The portable kit records access classes but cannot reproduce source-system permissions.
Institutional memory includes dissent
Overrides, rejected options, exceptions, corrections, and rationale remain available so later work can understand what happened and why.
Logical reference design
Eight layers connect source authority to governed BA work.
The reference design shows how an organization can extend the portable kit into a more integrated implementation while keeping evidence discovery, architecture identity, relationship traversal, agent reasoning, practitioner judgment, and institutional control distinct.
Governed source systems
Architecture platforms, portfolio systems, policies, decisions, approved documents, operational references, and other authoritative records remain identifiable and permission controlled.
Identity and provenance
Canonical identifiers, definitions, types, status, effective dates, ownership, confidence, aliases, and source evidence are resolved before a claim is trusted.
Architecture knowledge graph
A property graph holds durable, typed relationships among business and enterprise concepts so impact, dependency, overlap, and ownership paths can be traversed.
Evidence index
A semantic index helps find relevant source passages and candidate evidence. It supports discovery, but it does not become the authority for enterprise facts.
Question-aware retrieval
The engine classifies the question, selects relevant domains, applies access rules, traverses the graph, ranks evidence, and limits the response to useful scope.
Architecture reasoning
The agent proposes structures, comparisons, paths, findings, and challenges while preserving the distinction among evidence, interpretation, and recommendation.
Business Architecture Workbench
The business architect reviews proposals, investigates the architecture, records judgment, and composes the work product required by the decision context.
Workflow, identity, and audit
Permissions, reviews, acknowledgements, overrides, proposed changes, exceptions, evaluations, and human dispositions remain enforceable and inspectable.
Architecture Workspace
Model enough enterprise meaning to support the question.
SCALE does not require every domain to be complete. It requires stable identity, typed relationships, ownership, temporal status, access rules, and evidence for the questions it is expected to support.
Intent and value
Drivers, assessments, objectives, outcomes, stakeholders, value propositions, products, services, measures, and risks.
Operating architecture
Capabilities, value streams and stages, business objects, processes where needed, policies, rules, organization, roles, and decision rights.
Change and enablement
Initiatives, epics, requirements, decisions, applications, integrations, technology references, current states, target states, and transition dependencies.
Evidence and governance
Source artifacts, definitions, owners, status, versions, effective dates, confidence, classifications, reviews, exceptions, and change history.
Minimum element: stable identifier, type, name, definition, status, owner or steward, source, effective date, review date, confidence, access classification, version, and change history.
Minimum relationship: typed direction, source and target identifiers, status, evidence, confidence, owner, effective dates, and any scenario or scope qualifier.
First analysis
Begin with one decision. Continue with a governed architecture.
A bounded first analysis is an adoption path, not a temporary product identity. It produces a useful work product and establishes architecture that can be maintained and reused.
Name the business question
Identify the decision owner, audience, scope, time horizon, value at stake, evidence needs, and sufficiency criteria.
Assemble the evidence boundary
Link only the authorized sources, existing architecture, models, decisions, and owners needed for the question.
Build the relevant baseline
Propose and review the necessary concepts, relationships, status, provenance, conflicts, gaps, and ownership.
Produce the decision view
Trace implications, test assumptions, document confidence, compare options, and create the work product the audience requires.
Carry accepted knowledge forward
Retain validated relationships, evidence, corrections, decisions, review dates, and unresolved questions in the workspace.
Operating measures
Measure decision value, trust, reuse, and maintainability.
Targets should be calibrated to the organization, but the service should not confuse usage with value or model fluency with architecture quality.
| Measure | Operating expectation | Why it exists |
|---|---|---|
| Material claim traceability | Every enterprise fact has a governed record or cited source | Prevents fluent claims from outrunning evidence |
| Unsupported claim rate | Zero in accepted work products | Establishes a nonnegotiable trust boundary |
| Architecture review quality | Material proposals receive a named human disposition and rationale | Prevents machine suggestions from becoming implicit truth |
| Conflict and gap detection | Stale, missing, and contradictory records remain visible and actionable | Tests whether uncertainty is preserved |
| Access-control fidelity | No answer, citation, inference, or export exceeds source permissions | Tests the entire evidence path |
| Decision effort | Measured reduction in the time required to assemble and explain governed evidence | Shows whether the service reduces real institutional work |
| Architecture reuse | Accepted knowledge supports additional questions without losing context or provenance | Tests whether architecture is accumulating |
| Maintenance closure | Material corrections and reviews have owners, dispositions, and service expectations | Tests whether stewardship can sustain the workspace |
| Decision support | Named decisions are improved, accelerated, reframed, resequenced, or stopped | Connects the capability to institutional outcomes |
Failure modes
The main risks are institutional before they are technical.
A stronger model cannot cure ambiguous ownership, poor source authority, hidden exceptions, misaligned incentives, or an organization unwilling to maintain what the workspace presents as knowledge.
Fluent search wrapper
The system summarizes documents but cannot traverse governed relationships or distinguish source authority from textual similarity.
Document-first entry
The interaction begins with files and prompts instead of a business question, decision context, or architecture purpose.
Repository laundering
Poor definitions, stale ownership, and missing provenance acquire false credibility because an AI interface presents them confidently.
Permission collapse
Retrieval or citation reveals restricted content, sensitive relationships, or facts inferred from evidence the user could not access directly.
Invisible overrides
People disregard or alter a finding without recording the disposition, preventing learning and weakening institutional memory.
Activity as success
The service reports questions, responses, users, or model novelty instead of traceability, reuse, effort, correction closure, and decisions improved.
Architecture without stewardship
New concepts and relationships accumulate faster than named owners can validate, maintain, retire, or correct them.
Governance without capacity
The operating model assigns review and maintenance obligations to roles with no funded time, authority, or incentive to perform them.
Implementation patterns
The platform follows the operating model.
Start with the portable kit or extend SCALE through a more integrated technical pattern. Identity, relationships, evidence, permissions, ownership, review, and maintenance determine whether any implementation can be trusted.
Portable agent kit
Load the SCALE instructions, schemas, workbook, and reference material into an approved AI project or conversation. Export the workspace to maintain continuity outside conversational memory.
Productivity ecosystem
Use structured collaboration content, enterprise identity and permissions, supported APIs, an orchestration layer, and governed actions without treating document search as the architecture.
EA platform extension
Use the architecture platform as the governed element and relationship source, then add evidence retrieval, analysis, workbench, and workflow capabilities through supported interfaces.
Graph-native stack
Use a property graph for architecture links, a document store and semantic index for evidence, a relational audit store, and a model-agnostic orchestration service.
Lightweight workspace
Use governed Markdown or JSON, stable identifiers, explicit relationship files, source links, access rules, review status, and a focused retrieval service before buying enterprise tooling.
Take SCALE with you
Download the kit and begin in your own AI.
The ZIP includes the complete operating instructions, setup guide, launch prompt, schemas, workbook, reference architecture, and Northstar example. No registration or hosted SCALE service is required.
Have a use case or a design challenge?
Send the sanitized business question, repository pattern, or governance problem. Do not send proprietary architecture content.
Email a SCALE use case