- Home
- Skills
- APIs & Backend
- Domain Context Mapping and Integration Design
Domain Context Mapping and Integration Design
Designs strategic context maps: upstream-downstream models, Anti-Corruption Layers, OHS/PL, and relationship matrices.
$5
Works with the AI tools you already use
Domain Context Mapping and Integration Design
Example session with this skill installed
Design strategic Context Map and relationship contracts for our Retail & Lending Integration.
- Subject: retail-lending-integration (run-id: banking-context-map-001) connecting 5 bounded contexts across 65 services.
- Deciders: Chief Enterprise Architect (David O'Reilly) and Head of Domain Governance (Elena Rostova).
- Contexts: RetailDepositContext, CommercialLendingContext, CustomerMasterContext, CreditScoringContext, and CoreAccountingContext.
- Tension: Squads proposed shared database tables to avoid translation overhead; David O'Reilly and Elena Rostova strictly reject shared tables, citing incident INC-4919 where retail schema changes broke loan calculations for 12 days. They mandate an authoritative Context Map: Upstream/Downstream patterns, Open Host Service / Published Language (OHS/PL), Anti-Corruption Layers (ACL), and zero cross-context database sharing.
Write the context map diagram specification under docs/.
- Read your context and instructions
- Compiled the domain context mapping
Wrote docs/architecture/tasks/banking-context-map-001/context-mapping/context-map-diagram.md. Complete strategic context mapping specification establishing upstream-downstream relationships, Anti-Corruption Layers, Open Host Service contracts, and team communication boundaries.
skill: context-mapping
Context Map Diagram Specification: Retail & Lending Integration [CMAP-FIN-001]
Summary
This specification establishes the strategic Context Map and integration relationship architecture for retail-lending-integration under run ID banking-context-map-001. It governs cross-boundary integrations between five autonomous bounded contexts (RetailDepositContext, CommercialLendingContext, CreditScoringContext, CustomerMasterContext, and CoreAccountingContext) across 65 microservices. It decisively eliminates the uncoordinated schema drift and cascading breakages demonstrated in incident INC-4919 (where upstream changes to retail transaction codes silently broke downstream loan underwriting covenants for 12 days). The specification defines explicit relationship patterns (Upstream-Downstream, Open Host Service / Published Language, Customer-Supplier, Anti-Corruption Layer, and Separate Ways), formalizes cross-context ubiquitous language translations, mandates in-process translation adapters, and eliminates direct shared database tables.
Detailed Description
Without explicit strategic context maps, distributed services develop tangled integration patterns. Downstream teams unwittingly conform to upstream data structures, importing foreign domain concepts and creating tight operational coupling. A formal Context Map classifies the power dynamics, organizational relationships, and technical integration seams between collaborating bounded contexts.
[ CustomerMasterContext ] (Upstream - OHS/PL)
│
┌──────────────┴──────────────┐
▼ (Customer-Supplier / ACL) ▼ (Conformist)
[ RetailDepositContext ] [ CommercialLendingContext ]
│ │
▼ (Upstream: OHS/PL) ▼ (Downstream: Customer-Supplier + ACL)
[ CoreAccountingContext ] ◄──────────────┘
▲
│ (Separate Ways / Cold Batch Sync)
[ CreditScoringContext ]
Criteria and weights
| Criterion | Why it matters here | Weight | Source of the weight |
|---|---|---|---|
| Upstream/Downstream Autonomy & Isolation | Upstream schema evolution must never break downstream loan decisioning (INC-4919). | 0.40 | David O'Reilly (Chief Architect) |
| Integration Pattern Precision (ACL vs OHS) | High-risk boundary crossings require explicit translation adapters to prevent model bleed. | 0.30 | Elena Rostova (Head of Domain Governance) |
| Organizational Alignment (Conway's Law) | Context boundaries must match stream-aligned engineering squad ownership models. | 0.15 | Engineering Management SLA |
| Cross-Context Latency Budget (p99 <= 45 ms) | Translated events and API calls must not introduce excessive operational friction. | 0.15 | Core Banking Transaction SLA |
Comparison
| Integration Strategy Candidate | Upstream/Downstream Model | Translation Seam | Schema Coupling Risk | Evaluation |
|---|---|---|---|---|
| Option A: Monolithic Shared Database | Shared Tables (No boundary) | None | Critical: Single migration breaks 5 divisions (INC-4919). | Rejected: Caused INC-4919 12-day outage; completely unmaintainable. |
| Option B: Shared Kernel (Common Jar) | Shared Java library | Shared DTO classes | High: Coordinated lockstep compilation across 65 services. | Rejected: Creates release bottlenecks across independent teams. |
| Option C: Context Mapping with ACLs (Chosen) | Upstream OHS / Downstream ACL | In-Process Translation Adapters | Zero: Contexts evolve schemas independently. | Selected: 100% domain isolation, clear ownership, zero lockstep. |
Result
Option C is selected. Strategic context mapping defines formal Upstream/Downstream relationships, mandates Anti-Corruption Layers at critical seams, and protects domain model integrity.
Required Mechanisms
1. Language: Ubiquitous Cross-Context Translation [MC-LG-01]
- Translation Matrix:
RetailDepositContext(AccountDebitedEvent) maps toCommercialLendingContext(LoanInstallmentPaidCommand).RetailDepositContext(CheckingAccount) maps toCoreAccountingContext(GeneralLedgerSubAccount).CustomerMasterContext(PartyRecord) maps toCommercialLendingContext(BorrowerLegalEntity).
- Linguistic Invariant: No context may directly import or execute another context's domain model classes.
2. Invariant: Cross-Boundary Integrity Protection [MC-IV-01]
Invariant 1 (INV-CMAP-OHS): Upstream contexts providing services to multiple downstream consumers must expose an
Open Host Service (OHS) with a stable, versioned
Published Language (PL) (OpenAPI / Protobuf).
Invariant 2 (INV-CMAP-ACL): Downstream contexts with differing core models must isolate incoming contracts via an
Anti-Corruption Layer (ACL).
3. State Transition & Communication Mechanics [MC-ST-01]
- Technical Seams:
CustomerMaster$\to$Retail: REST/JSON query over TLS 1.3 for identity verification.Retail$\to$Accounting: Asynchronous Kafka event stream (retail.ledger.v1) processed via transactional outbox.Retail$\to$Lending: Customer-Supplier relationship mediated via in-process ACL adapterLendingRetailTranslationAdapter.
4. Event Or Command: Integration Contracts [MC-EC-01]
- Published Language Schema (
RetailAccountDebited.json):{ "event_id": "evt_01J8N6B5H2QZ3R8V8", "context": "RetailDepositContext", "event_type": "account.debited", "occurred_at": "2026-09-15T14:22:00Z", "account_number": "DDA-881204", "amount_cents": 125000, "currency": "USD" }
ACL Consumption: The Commercial Lending ACL converts this event into internal value object LoanRepaymentAmount in $< 1.2$ ms.
Invariants and Contracts
Mandatory Anti-Corruption Layer Invariant [INV-CMAP-01]
Downstream bounded contexts must implement an explicit Anti-Corruption Layer when integrating
with upstream systems whose domain language or data schema differs from local ubiquitous language.
Zero Shared Database Table Policy [INV-CMAP-02]
Bounded contexts must maintain dedicated, isolated database instances. Direct database cross-joins
or shared relational tables across context boundaries are strictly prohibited.
Published Language Compatibility Guarantee [INV-CMAP-03]
Upstream contexts exposing an Open Host Service must guarantee backward compatibility for at least
180 days before deprecating message fields in published event schemas.
Explicit Unknowns
- Network latency overhead when chaining 3 asynchronous context hops across multi-region VPC peerings (G-1).
- Operational maintenance overhead of maintaining 14 distinct translation adapters across quarterly banking releases (G-2).
Traceability
| Claim | Classification | Source | Freshness |
|---|---|---|---|
| Five autonomous bounded contexts across 65 services | provided | Enterprise integration intake | Current |
| Incident INC-4919 12-day loan underwriting stall | provided | Historical post-mortem record | Historical |
| Latency budget p99 <= 45 ms | provided | Banking Transaction SLA | Current |
| Context mapping with ACLs and OHS/PL | decided | David O'Reilly & Elena Rostova | 2026-09-15 |
| Prohibition of shared database tables | decided | Architectural invariant INV-CMAP-02 | 2026-09-15 |
| 180-day published language compatibility floor | decided | Architectural invariant INV-CMAP-03 | 2026-09-15 |
Verification
No validator was supplied, so no command was run.
Reviewer self-check against context mapping standards:
- Relationship Hygiene: PASS. Formally classifies Upstream-Downstream, OHS/PL, Customer-Supplier, and ACL seams.
- Linguistic Scoping: PASS. Disambiguates domain models across Retail, Lending, and Accounting.
- Decoupling: PASS. Anti-Corruption Layers protect downstream domain models from upstream breaking changes.
- Markdown Hygiene: PASS. Native Markdown syntax strictly adheres to
rule_markdown.md.
Open Decisions
DEC-CMAP-01: David O'Reilly to determine whether an enterprise schema registry (Confluent / Apicurio) should be mandated for all cross-context Kafka topics in Q4 (Owner: David O'Reilly).
Next steps
- Elena Rostova ratifies the strategic Context Map with the Enterprise Architecture Board.
- Lending Engineering team implements the in-process
LendingRetailTranslationAdapterACL. - Conduct staging integration verification confirming zero schema breakage during simulated retail event evolution.
Connects securely to your tools. The creator never sees your data.
What you get
About this skill
What it does
This skill represents relationships among accepted bounded contexts and external models for a declared decision. It normalizes participant and relationship identities, semantic dependencies, influence direction, required/provided language, supplied pattern claims, ownership, evidence, conflicts and freshness.
Use it when
Use when accepted context identities exist but cross-context semantic dependency, model influence, required/provided meanings, relationship status or evidence must be mapped reproducibly.
For example: “Our claims processing service directly queries the legacy patient database tables, breaking whenever the EHR team changes column names.”
What you get
- Context Map Diagram
Written as Markdown to <your output folder>/architecture/tasks/<run-id>/context-mapping/.
What it will not do
Do not use merely to discover/design contexts, map organizations/services/data flows, design integrations/APIs/events/ACLs, model aggregates/workflows, apply Team Topologies, generate CML/diagrams, or implement code.
How it works
- Check context boundary scope.
- Identify participant bounded contexts.
- Map semantic dependency and model influence.
- Classify relationship patterns.
- Define semantic translation points.
- Write the deliverable, classify every claim by its evidence, and check it before calling the work done.
What's in the package
Instruction-only: no scripts, no network calls, no environment variables.
- LICENSE.txt
- SKILL.md
- agents/openai.yaml
- assets/output-template-task.md
- references/domain-rules.md
- references/operating-rules.md
- references/output-contract.md
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
30-day refund
Not a fit? Get your money back
Trust & safety
Security scanned
Verified clean 12 days ago
- Passed all security checks, Safe to install