Use n8n to draft YouTube scripts from trend signals

Ben Utting built a content-research pipeline for a client running YouTube channels across several niches. The client was manually scanning Reddit, competitor uploads and Google Trends, then writing scripts and checking whether an idea had already been covered. Utting replaced that preparation work with scheduled n8n sub-workflows.

The system gathers research signals, removes repeated topics, scores ideas and prepares search metadata, scripts and quality checks before synchronizing with Notion. Four AI agents handle the reasoning stages. A niche configuration identifies subreddits, channels, keywords and scoring preferences; a PostgreSQL backend supports the workflow.

The client reviews scored ideas and first drafts in Notion, chooses what to film and edits the scripts before production. Utting reports reduced preparation effort and fewer accidental repeats. The code and prompts are linked as an open-source artifact, with prompts kept in separate Markdown files for tuning. This account documents draft preparation and a human filming decision. It does not describe automatic video publication or prove that the system's topic scores predict audience success.

Prompt
Suggested prompt (written by Agents In Use, not the author's)

Source: dev.to · Ben Utting