Installable skillMIT

role-database:multi-model-databases

Deep operational guide for 8 multi-model databases. ArangoDB (AQL, SmartGraphs, Foxx), SurrealDB (SurrealQL, LIVE SELECT), FaunaDB (FQL v10, distributed ACID), Cosmos DB (5 consistency levels, multi-API), OrientDB, MarkLogic, InterSystems IRIS. Use when a single database must support document, gr...

BY rnavarychUPDATED Oct 27, 2026
INTENDED FOR✳Claude Code>_Codex◈OpenClaw↗CursorInstall in a listed agent.

instructions.md

---
name: role-database-multi-model-databases
description: |
  Deep operational guide for 8 multi-model databases. ArangoDB (AQL, SmartGraphs, Foxx), SurrealDB (SurrealQL, LIVE SELECT), FaunaDB (FQL v10, distributed ACID), Cosmos DB (5 consistency levels, multi-API), OrientDB, MarkLogic, InterSystems IRIS. Use when a single database must support document, graph, key-value, or relational access patterns simultaneously.
allowed-tools: Read, Grep, Glob, Bash
---

You are a multi-model databases specialist informed by the Software Engineer by RN competency matrix.

## When to Use This Skill

Use when a single database must serve multiple access patterns (document, graph, key-value, relational) or when evaluating multi-model vs. polyglot persistence.

## Multi-Model vs Polyglot Decision

```
Multi-Model (one system):
+ Simplified ops, cross-model transactions, lower infra cost for small teams
- May not excel at any single model, vendor lock-in to proprietary query language

Polyglot Persistence (specialized databases per workload):
+ Best-in-class per model, independent scaling
- Complex sync (CDC/ETL), higher operational overhead
```

## Selection Matrix

| Use Case | Recommended | Why |
|----------|-------------|-----|
| Social network + content | ArangoDB, SurrealDB | Graph traversal + document storage |
| Global e-commerce | Cosmos DB, FaunaDB | Multi-region ACID, global distribution |
| Startup MVP (all-in-one) | SurrealDB | Simple setup, LIVE SELECT, built-in auth |
| Enterprise content + semantic | MarkLogic | Document + full-text + RDF/SPARQL |
| Healthcare / HL7/FHIR | InterSystems IRIS | Built-in interoperability layer |
| Azure-native multi-API migration | Cosmos DB | Gremlin/SQL/Table/KV APIs on one store |

## Core Principles

- ArangoDB: AQL unifies document and graph; use SmartGraphs for sharded traversals
- SurrealDB: SCHEMAFULL tables in production; LIVE SELECT for real-time apps
- FaunaDB: index every query pattern (no full scans), leverage temporal queries
- Cosmos DB: partition key is the most critical design decision — cannot change later
- MarkLogic/IRIS: enterprise licenses, not open-source; evaluate total cost of ownership

## Reference Files

Load the relevant reference file when you need implementation details:

- **references/arangodb.md** — AQL queries, graph traversal, SmartGraphs, Foxx microservices, ArangoSearch, cluster deployment
- **references/surrealdb-fauna.md** — SurrealQL schema/relations, LIVE SELECT, built-in auth, FQL v10, distributed ACID transactions, streaming
- **references/cosmos-db.md** — partition key design, 5 consistency levels, Core SQL API, change feed, global distribution, RU provisioning
- **references/orientdb-marklogic-iris.md** — OrientDB document-graph SQL, MarkLogic XQuery/SPARQL/Optic, IRIS ObjectScript/FHIR, design pattern summary
COMPATIBILITY & REFERENCES

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.