AI Content Labs hides the intermediate draft. What can the webhook still receive?

When chaining content nodes in AI Content Labs, hiding output only controls the final presentation. Teams using webhooks must also check each node’s send settings and verify the receiving end with fictional data. Out of sight does not mean the data has stopped flowing.

Source: docs.aicontentlabs.com
November 5, 2026
4 min

Content teams often summarize product information, then have downstream nodes write emails and social copy without delivering the intermediate summary alongside them. AI Content Labs’ Hide Node Output can tidy the final presentation, but once a webhook is connected, you must separately decide which node results may go to an external system. The two options govern different exits; a single “hide” switch cannot replace checking them.

The intermediate draft still feeds later generation

Prompt nodes in AI Content Labs can reference earlier nodes through variables. Official documentation says a connection can be established automatically when a variable name matches an upstream node’s title, or manually when it does not. The official flow-creation example uses an executive summary as a reference for later sections while keeping it out of the final report.

This suits situations where you want the preparation to stay within the flow and only show the finished copy. Suppose a product-copy flow first generates a “selling-points summary,” then uses it to produce a “customer email.” Hiding the former makes the final result cleaner, but does not remove its influence on the latter. If the summary contains an incorrect feature description, the customer email may repeat it.

During debugging, therefore, keep the intermediate draft visible, check how the source material becomes a summary and how that summary enters the finished copy, and hide it only after verification. This is a proposed verification method based on documented behavior, not a flow actually run for this article. For a simple rewrite of one passage, a clear single prompt is often easier to inspect than several nodes. Do not add intermediate drafts just to make the flowchart look complete.

Check what the webhook receives separately

In a Prompt node’s output settings, Hide Node Output hides the response from the final result. A separate option, Do Not Send to Webhook, prevents that node’s result from being sent to an enabled webhook. Given these distinct responsibilities, you cannot infer that hiding a node automatically excludes it from external delivery.

In the product-copy example, you could disable webhook delivery for the “selling-points summary” while keeping it enabled for the “customer email” that needs to be handed over. This still is not data sanitization: if the downstream email repeats internal notes from the summary, blocking delivery of the summary node will not automatically remove that same content from the email.

Before going live, check the receiving end using fictional product information. Put an easily recognizable test marker containing no real sensitive information into the intermediate summary, inspect both the final presentation and the fields actually received by the webhook, then check whether the downstream copy carries the marker forward. This answers both “which node sent what?” and “did downstream content carry it along?”, which is more useful than looking only at the flow canvas. This article has not verified the webhook’s actual request format; use the data received in your own run as the basis for checking the receiving end.

Hiding, external delivery, and history have different boundaries

Another button is easy to misunderstand. The official results-panel documentation says the trash icon deletes the generated result from the interface, while the record remains in the account’s history. Removing content from the current screen should therefore not be treated as permanently deleting it from the account.

Flow visibility is another separate matter. The creation page describes Private as available only to you and Public as available to the community. That describes who can use the flow; it does not establish that a given input will not be sent to a model or another connected service. For customer information, decide what may be entered under your team’s data-use requirements before designing the nodes, rather than relying on hiding to fix the problem afterward.

The value of this feature is in delivering only the necessary results from a complex generation process. A reliable configuration checks separately that intermediate content is correct, the final presentation is concise, and external recipients receive only permitted data. Hide Node Output addresses the presentation part. The other two still require explicit checks.

Content note:This article is based on publicly available information and is intended as a reference for evaluating tools. Features, pricing, and related policies may change at any time; please refer to the latest official information.

Original source:docs.aicontentlabs.com