---
title: "Cockpit Integration"
description: "Connect AI agents to Cockpit CMS through the v2 Content API. Agents read and write content items, singletons, and assets, query with MongoDB-style filters, and stage or remove entries as part of an editorial pipeline."
url: https://flowrunner.ai/integrations/cockpit
date_modified: 2026-08-07T17:20:23-07:00
---

# Cockpit

[CMS & Content](https://flowrunner.ai/integrations/category/cms-content)

Connect AI agents to Cockpit CMS through the v2 Content API. Agents read and write content items, singletons, and assets, query with MongoDB-style filters, and stage or remove entries as part of an editorial pipeline.

[Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 10 actions · API key · available

[Cockpit website](https://getcockpit.com/) · [Platform Documentation](https://getcockpit.com/documentation/api/content) · Capability data verified 2026-07-14

1.  A content request arrives from an upstream system or channel
2.  Agent reads the source data and maps it to the target model
3.  Agent checks the required fields for that model are present
4.  Agent writes the entry with Save Content Item
5.  Agent confirms the saved item returned an id
6.  A person approves any permanent delete before it runs

## What This Integration Enables

An agent can list items from a content model with MongoDB-style filtering, sorting, field projection, pagination, and localization, get a single item by id or filter, and read a tree-structured model as a nested tree. It reads singletons such as site settings or a homepage to drive downstream automations, and counts items matching a filter to build reports or gate a conditional flow.

On the write side, Save Content Item both creates and updates, distinguished by whether you pass an item id, and Update Content Item is a convenience wrapper for updates. The agent can also browse assets in the asset manager and read a single asset's metadata. Delete Content Item is permanent, so the agent treats it as the one step that always needs a person before it runs.

### Without FlowRunner

**Manual content entry**: Someone retypes external data into Cockpit collections by hand

**No safety on deletes**: A content item can be removed with no second look, and it is gone

**Reporting is manual**: Item counts are gathered by scrolling through collections

### With FlowRunner

**Entries written by agent**: The agent maps source data and saves it to the right model

**Deletes are reviewed**: A person approves before the agent removes an item permanently

**Counts on demand**: The agent counts items matching a filter for reports and checks

## Use Case Scenarios

### Sync Content Into Cockpit

The agent reads a document from [Google Drive](https://flowrunner.ai/integrations/google-drive-api), extracts the data, and calls Save Content Item to create a matching entry in the target model. Content that used to be retyped by hand lands in Cockpit already structured.

### Summarize and Enrich Entries

The agent uses Get Content Items to pull entries matching a filter, sends each one to a model to generate a summary, and writes the result back with Update Content Item. A backlog of bare entries becomes a set of enriched, summarized records without anyone editing them one by one.

### Draft From a Channel Request

When a content request comes in through [Slack](https://flowrunner.ai/integrations/slack), the agent creates a new entry in the target model with Save Content Item and replies in the channel with the new item id. The requester gets a confirmation and a reference instead of waiting for someone to key the entry in later.

## Human-in-Loop Highlight

Reading, counting, creating, and updating content are safe to automate, so the agent handles them without stopping. Deleting is different, because Delete Content Item is permanent and cannot be undone. So when a flow reaches a delete step, the agent stops. It routes the request to a person as a [human-in-the-loop](https://flowrunner.ai/concepts/human-in-the-loop) step, showing the model, the item id, and a snapshot of the item's fields, and asks: "Permanently delete this content item, or keep it?" On approval the agent calls Delete Content Item. On hold, nothing is removed. The AI handles every reversible change on its own; the one action it cannot take back gets a person.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

## Agent Capabilities

10 actions

### Content Items

7

-   **Get Content Items** Lists entries from a content model with MongoDB-style filtering, sorting, field projection, pagination, linked-content population, and localization.
-   **Get Content Item** Retrieves a single entry by id, or the first item matching a MongoDB-style filter, with field projection, population, and localization.
-   **Save Content Item** Creates or updates an entry in a content model. Omit the item id to create a new item; provide it to update the matching entry.
-   **Update Content Item** Updates an existing entry by id with the field values you set. A convenience over Save Content Item for updates, where the item id is required.
-   **Delete Content Item** Deletes an entry from a content model by id. This is permanent and cannot be undone.
-   **Get Content Tree** Retrieves entries of a tree-structured content model as a nested tree, with child items attached to their parents.
-   **Count Content Items** Returns the number of items in a content model that match an optional MongoDB-style filter, so use a filter to narrow large collections.

### Singletons

1

-   **Get Singleton** Retrieves the content of a Cockpit singleton by name, such as site settings or a homepage, returning its field values.

### Assets

2

-   **List Assets** Lists assets from the asset manager with MongoDB-style filtering, sorting, and pagination, returning metadata such as id, path, title, mime type, and size.
-   **Get Asset** Retrieves metadata for a single asset by id, including its path, title, mime type, size, and dimensions.

## Frequently Asked Questions

### What can FlowRunner do with Cockpit?

FlowRunner agents can run Get Content Items, Get Content Item, and Get Singleton in Cockpit, plus 7 more actions.

### Does connecting Cockpit to FlowRunner require OAuth?

No. Cockpit connects to FlowRunner with an API key, no OAuth flow required.

### Can Cockpit trigger a FlowRunner workflow automatically?

Cockpit doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

**Work at Cockpit?** This integration exposes Cockpit to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current.

---
Markdown version of https://flowrunner.ai/integrations/cockpit. Site index: https://flowrunner.ai/llms.txt
