Overview

The YouTube Brain is a source-cited knowledge and decision layer for YouTube, spanning three pillars: creator growth and strategy, the developer API, and paid advertising. Its job is to make repeatable, defensible decisions with every claim traceable to an official or primary source. It serves creators, channel managers, editors, strategists, tool builders, and advertisers. See YouTube Brain Home to start reading.

Scope

The brain now covers the whole YouTube surface a practitioner touches:

It also carries patterns to apply, antipatterns to avoid, and flows to run, all listed in index.

Sourcing

Every claim is researched from official and primary sources. The result is 156 structured sources in references/source-ledger.json and 143 dated citations in research-pack-2026-06-23, the majority from YouTube Help, the YouTube Official Blog, How YouTube Works, developers.google.com, and Google Ads Help. Domain adapters in scripts/ and youtube_brain/adapters.py turn a Studio export into a sourced channel-health scorecard.

How it stays current

YouTube’s rules move fast, so the brain refreshes on a cadence: monthly for Help Center, Studio, and Google Ads pages, on changelog for recommendation, monetization, API, and ad-campaign changes, and quarterly for practitioner benchmarks. The discipline lives in Source Intake and Refresh.

Safety posture

  • No credentials, OAuth tokens, cookies, API keys, or private audience data ever enter the repo.
  • No growth, revenue, monetization, API, or ad claim ships without a current official or primary source.
  • No account mutation; the brain is advisory and read only toward any real channel, project, or ad account.
  • No sub for sub, bought engagement, fake API traffic, or policy-violating tactics; see Bought Engagement and Sub4Sub and API Policies and Limits.
  • No irreversible recommendation without owner, confidence, source, and a rollback note.

Sources