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

> ## Agent Instructions
> Cognite Data Fusion (CDF) is an industrial DataOps platform. English is the canonical locale unless the reader asks for another.
> The documentation catalog is https://docs.cognite.com/llms.txt. Subgroup indexes live under https://docs.cognite.com/_llms/. Each published page has a Markdown twin at the same path with a .md suffix.
> REST APIs: calendar version 20230101 is stable. Preview versions are 20230101-beta and 20230101-alpha, for clients already on a preview API. Concept pages are under /api-reference/concepts/<version>/. Versioned requests use the Cdf-Version header. The default SDK is the Python SDK (cognite-sdk). JavaScript and TypeScript projects use the JavaScript SDK.
> Topic hubs: platform overview /cdf/index.md; data modeling /cdf/dm/index.md; data integration /cdf/integration/index.md; access /cdf/access/index.md; Cognite Toolkit /cdf/deploy/cdf_toolkit/index.md; REST quickstart /dev/quickstart.md; Python SDK /dev/sdks/python/index.md; Atlas AI /cdf/atlas_ai/concepts/index.md; Flows apps /cdf/flows/index.md; Aura UI /cdf/aura/index.md; MCP and IDE setup /dev/guides/ide_ai_integration.md.

# Integrations resource

> Register, update, and delete integrations so you can monitor extractors and data pipelines in Cognite Data Fusion (CDF).

Integrations represent extractors and other data pipelines that CDF monitors. Each integration tracks the tasks it runs, the errors it reports, and whether the extractor is running. You can create, list, retrieve, update, and delete integrations.

An integration is a record of an extractor or other process that sends data to CDF. It identifies the extractor type and holds the configuration, task history, and error history for that data pipeline. Tasks, actions, configuration, and errors are all scoped to an integration.

## Managing integrations

You can create integrations when you set up a new extractor, and update them as the extractor's details change. Use list and retrieve to find integrations by ID and inspect their current state. Delete an integration when the extractor it represents is decommissioned.

<Note>
  Deleting an integration removes it from monitoring. After you delete it, you cannot retrieve its task and error history.
</Note>

## Key operations

* **Create**: Register a new integration for an extractor or data pipeline.
* **List and retrieve**: Find integrations and inspect their details.
* **Update**: Modify integration metadata as the extractor changes.
* **Delete**: Remove an integration that is no longer in use.
* **Startup**: Report the extractor's type, version, and configured tasks when it starts.
* **Check in**: Report task and error updates periodically, and receive the active configuration revision and pending actions.

Integrations are the anchor that task history, actions, configuration, and errors are all reported against.

## Extractor startup and check-in

Extractors report that they are running directly to the integration record. On startup, an extractor reports its type, version, and configured tasks. It also closes any tasks still marked as running from a previous session. After that, it checks in periodically. Check-in reports task starts, stops, and errors, and returns the latest configuration revision and any pending actions.

<Note>
  Startup and check-in require the `integrationsAcl:USE` capability, a separate scope from the `READ` and `WRITE` capabilities used by the other integration operations. Only the extractor's own credentials should have `USE` for an integration. Reporting status on behalf of an integration you do not operate overwrites its real running status, task history, and error history with incorrect data.
</Note>