> ## Documentation Index
> Fetch the complete documentation index at: https://docs.insight.nobly.dk/llms.txt
> Use this file to discover all available pages before exploring further.

# How the system fits together

> Understand the six Nobly Insight components and the configuration boundaries between identity, content, scripts, and workflow automation.

Nobly Insight comprises six cooperating components. The product's configuration spans these components, so identify which one owns a behavior before changing it.

| Component           | Role                                                                                | Where to work                                                    |
| ------------------- | ----------------------------------------------------------------------------------- | ---------------------------------------------------------------- |
| **Client**          | The browser interface for documents, Caseflow, reporting, and administration        | Insight's pages and Admin settings                               |
| **API**             | The public contract for content, configuration, permissions, and integrations       | Client administration screens or the environment's API reference |
| **Script Engine**   | Python extensions for indexing, Caseflow, workflow actions, and document events     | Admin settings → Script Engine                                   |
| **Workflow Engine** | Document, Caseflow, and system workflows with actions, tasks, timers, and schedules | Admin settings → Workflow                                        |
| **ContentFlow**     | Document ingestion and transformation pipelines with mapping and processing traces  | The separate ContentFlow application for your environment        |
| **Identity**        | Sign-in and application identities, including federation and service clients        | Your identity provider and Insight's Access administration       |

## Follow one document through the system

```mermaid theme={null}
flowchart LR
  Person[Person] --> Identity[Identity: sign-in]
  Person --> Client[Client: find and manage]
  Client --> API[API: content and permissions]
  Intake[Incoming files] --> CF[ContentFlow: transform and map]
  CF --> API
  API --> WE[Workflow Engine: tasks and actions]
  WE --> SE[Script Engine: focused extensions]
  Client --> SE
  SE --> API
```

For a contract, ContentFlow might map incoming metadata and upload it through the API. A configured workflow can then request a human review. The person completes that task in the Client, while a script supplies a specific lookup or decision. These are possible configured connections, not steps that every document automatically follows.

## Select the right automation

Use a **script** for a focused operation, such as looking up indexing metadata or reacting to a form action. Use a **Workflow Engine definition** when a process has state over time, human decisions, timers, or scheduled batches. Use **ContentFlow** to process incoming content through conversion, mapping, and integration steps.

These systems have separate definitions and lifecycles. Publishing a Python script does not publish a workflow. Restoring a ContentFlow version changes its draft, while rolling back a script changes its published version. Read the lifecycle for the component you are changing.

## Access has several layers

Identity establishes who is calling. User-group membership and application permissions govern capabilities; document-type rights and record restrictions govern content access. A system integration uses its own service user, while a background workflow or event hook can use a configured service identity.

Granting access to a configuration screen is not the same as granting its resulting automation access to every document. Verify the identity used during execution as well as the identity used to configure it.

## Configure and verify in a test environment

1. Define the intended business behavior and the component responsible for it.
2. Confirm feature availability and the permissions needed to configure and execute it.
3. Configure with representative test metadata and sample content.
4. Validate or preview where supported, then execute a controlled test and inspect the result and logs.
5. Publish or enable through that component's lifecycle and verify the version and environment.

Names, IDs, mappings, credentials, and available features can differ across environments. Exporting a definition alone does not transfer all its dependencies.

## Where to read next

<Card title="Features overview" icon="list-check" href="/features/introduction" horizontal>
  What Nobly Insight does — document management, smart search, case management, workflow, AI, extensibility, and governance — at a glance.
</Card>

<Card title="Document configuration" icon="tags" href="/configuration/introduction" horizontal>
  Start with the document types and metadata that other features depend on.
</Card>