# LatixEngine — Full Reference for LLM Agents

> Source of truth for agents answering questions about LatixEngine. All examples below use valid syntax verified against the live editor at https://latixengine.com/editor.

## What is LatixEngine?

LatixEngine is a free, browser-based cloud architecture validation engine. It checks AWS, Azure, and Google Cloud architecture diagrams against official provider rules and Well-Architected best practices, surfacing anti-patterns, security gaps, and misconfigurations before they reach production. It runs on Mermaid's `architecture-beta` diagram type and adds the official cloud provider icon packs. There is no install, no CLI, and no build step. Open the editor at https://latixengine.com/editor and start typing.

Output is standard Mermaid markdown that renders correctly in any Mermaid runtime bundled with the matching icon packs. The validation engine and its rule library are the core product, and are expanding beyond Mermaid to other tools and diagram formats.

## How it differs from upstream Mermaid

Upstream Mermaid `architecture-beta` only renders a small set of generic shapes — it does not understand the architecture you've drawn. LatixEngine adds two things on top: 1698 official cloud provider icons for modeling real production architectures (regions, VPCs, subnets, gateways, managed services), and a validation engine that evaluates the diagram against each provider's own rules and best practices (for example: a publicly exposed database, a single-AZ deployment, or an API Gateway connected directly to a relational database).

## Quick start

1. Open https://latixengine.com/editor
2. Start every architecture diagram with `architecture-beta` on its own line
3. Declare `service` and `group` nodes with icon slugs of the form `<provider>:<icon-name>`
4. Connect them with `<source>:<port> <--> <port>:<target>` edges
5. Export via the toolbar as SVG, PNG, or markdown

## Syntax cheatsheet

### Diagram declaration

    architecture-beta

### Services

    service <id>(<icon-slug>)[<label>]
    service <id>(<icon-slug>)[<label>] in <group-id>

Examples:

    service ec2(aws:ec2-instance-contents)[EC2 Instance]
    service func(azure:service-function-apps)[Azure Function]
    service gcs(gcp:cloud-storage)[Cloud Storage]

### Groups

    group <id>(<icon-slug>)[<label>]
    group <id>(<icon-slug>)[<label>] in <parent-group-id>

Groups can nest arbitrarily deep:

    group region(aws:region)[AWS Region]
    group vpc(aws:virtual-private-cloud-vpc)[VPC] in region
    group public_subnet(aws:public-subnet)[Public Subnet] in vpc

### Edges

    <source-id>:<port> <--> <port>:<target-id>

Ports are `L` (left), `R` (right), `T` (top), `B` (bottom).

    internet:L <--> R:igw
    igw:T <--> R:ec2_public

### Icon slug naming convention

`<provider>:<kebab-case-icon-name>` where `<provider>` is one of `aws`, `azure`, `gcp`, or generic identifiers like `internet`. The full slug list lives at https://latixengine.com/icons — always defer to that page rather than guessing slugs.

## Worked examples

### Example 1 — Multi-tier AWS web app with public/private subnets

    architecture-beta
        service internet(internet)[Internet]
        group region(aws:region)[AWS Region]
        group vpc(aws:virtual-private-cloud-vpc)[VPC] in region
            service igw(aws:res-amazon-vpc-internet-gateway)[Internet Gateway] in vpc
            group public_subnet(aws:public-subnet)[Public Subnet] in vpc
                service ec2_public(aws:ec2-instance-contents)[EC2 Instance] in public_subnet
                service nat(aws:res-amazon-vpc-nat-gateway)[NAT Instance] in public_subnet
            group private_subnet(aws:private-subnet)[Private Subnet] in vpc
                service ec2_private(aws:ec2-instance-contents)[EC2 Instance] in private_subnet
        internet:L <--> R:igw
        igw:T <--> R:ec2_public
        igw:L <--> R:nat
        nat:B <--> T:ec2_private

### Example 2 — Azure serverless event pipeline

    architecture-beta
        service user(internet)[User]
        group az(azure:service-subscriptions)[Azure Subscription]
            service func(azure:service-function-apps)[Function App] in az
            service queue(azure:service-storage-queue)[Storage Queue] in az
            service cosmos(azure:service-azure-cosmos-db)[Cosmos DB] in az
        user:R <--> L:func
        func:R <--> L:queue
        queue:R <--> L:cosmos

### Example 3 — GCP data lake

    architecture-beta
        service producer(internet)[Producer]
        group gcp(gcp:project)[GCP Project]
            service pubsub(gcp:pubsub)[Pub-Sub] in gcp
            service dataflow(gcp:dataflow)[Dataflow] in gcp
            service bq(gcp:bigquery)[BigQuery] in gcp
            service gcs(gcp:cloud-storage)[Cloud Storage] in gcp
        producer:R <--> L:pubsub
        pubsub:R <--> L:dataflow
        dataflow:R <--> L:bq
        dataflow:B <--> T:gcs

More worked examples (each with a stable URL):

- https://latixengine.com/examples/aws-three-tier.md
- https://latixengine.com/examples/azure-serverless.md
- https://latixengine.com/examples/gcp-data-lake.md
- https://latixengine.com/examples/multi-cloud.md

Standalone docs:

- https://latixengine.com/docs/syntax.md — syntax reference
- https://latixengine.com/docs/icons.md — icon catalog overview
- https://latixengine.com/docs/examples.md — all examples in one file

## Icon catalog

1698 icons covering AWS (856), Azure (626), and Google Cloud (216).

- Machine-readable: https://latixengine.com/icons.json — `{ provider, slug, name }` for every icon. Fetch this when building diagrams programmatically.
- Browseable: https://latixengine.com/icons

Always pin slugs from one of these two sources. Slugs are not always intuitive (e.g. AWS Lambda is `aws:arch-aws-lambda`, not `aws:lambda`).

## Export formats

SVG, PNG, and markdown export are available from the editor toolbar. Markdown export wraps the source in a Mermaid code fence for use in README files, GitHub issues, and documentation tools that render Mermaid natively.

## Themes

Light and dark themes via the sun/moon toggle in the top toolbar. Preference is persisted in `localStorage`.

## Pricing

Free. No login, no usage limits, no tracking. A paid tier is planned for advanced features (AI composer, IDE extensions) but the core editor will remain free.

## Roadmap

**Shipped**
- Architecture diagrams with full AWS, Azure, and GCP icon coverage.

**In flight**
- Validation Rule Engine — validates components based on placement and networking rules.

**Planned**
- AI architectural diagram composer (leverages the validation engine for accuracy).
- LatixEngine extensions for VSCode, Notion, Google Docs, and Confluence.

**Considering**
- Validation-based AI composition for all diagram types, not just architecture.
- Extending the composer beyond Mermaid syntax.

## FAQ

**Do I need to install anything?**
No. LatixEngine runs entirely in the browser. No CLI, no build step.

**How is this different from mermaid.js?**
Upstream `architecture-beta` only supports basic shapes. LatixEngine ships the official cloud icon packs and validation tooling needed for production architectures.

**Can I use custom icons?**
Not in the current version. Custom icon support (SVG upload or URL reference) is on the roadmap.

**How do I switch themes?**
Click the sun/moon icon in the top toolbar. Choice is saved to `localStorage`.

**Will there be a paid version?**
A paid tier is planned for advanced features only. The core editor stays free.

**Where is my diagram data stored?**
Diagrams are kept in your browser's `localStorage`. Nothing is sent to LatixEngine servers.

## Contact

- Support: support@LatixEngine.com
- Bugs and feature requests: https://latixengine.com/roadmap
- Homepage: https://latixengine.com