Founders, solo marketers, agencies and in-house teams can use this page to install a complete marketing department built from AI agents, then have it producing publishable work the same afternoon. Inside are 50 role-based agents arranged across 10 departments, a shared Brand Brain template with two filled examples, an operating guide, a directory of every agent, 5 connected workflows, 100 prompt starters and three quality control checklists.

<aside> ⚠️

Before using these prompts you should omit any PII (personal identifiable information) and check with your compliance team and/or your AI policy.

</aside>

<aside> 🥷

Want this team configured around your brand, your offers and your voice before you run a single prompt? Talk to UGC Ninja

</aside>

Every number on this page describes the structure of the system itself, not a client's measured results.

How the counts resolve

Every claim on this page is countable on the page itself, so nothing has to be taken on trust.

Claim How it resolves Where you check it
50 agents 10 departments with exactly 5 agents each, every agent carrying its own name, purpose, inputs, process, output format and paste-ready prompt. No agent is counted twice. The agent directory table, then the ten department sections below it
10 departments Research, Strategy, Positioning, Content, Social Media, Video, Creative, Distribution, Campaigns, Analytics. Each is a second level heading with five third level agent headings under it. The department headings running from Research through to Analytics
100 prompt starters 10 starters per department, numbered continuously from 1 to 100 with no gaps and no repeats. The prompt starters section near the end of the page
5 workflows Campaign creation, organic content, product launches, lead magnets and weekly reporting. Each names its agents in running order, the handoff between them and the usual failure point. The connected workflows section

The operating guide: install once, then run the team every week

The team is not software. It is a set of written roles you paste into Claude, plus one shared context file every role reads before it works. Install it once and the running cost is measured in minutes.

Installing the team in Claude

  1. Open Claude and create a new Project named after your brand. A Project keeps context in one place so you are not re-explaining the business in every chat.
  2. Copy the Brand Brain template from the next section of this page into a document, fill it in, and add it to the Project knowledge. This is the single most valuable hour you will spend on the whole system.
  3. Create one chat per agent you use regularly. Name the chat after the agent, for example "Positioning: Objection Handling Architect". Paste that agent's prompt in as the first message.
  4. For agents you use occasionally, keep one scratch chat and paste the prompt fresh each time.
  5. Re-read the Brand Brain every quarter. Stale context is the most common cause of output that feels almost right but is not usable.

How the Brand Brain feeds every agent

Every agent prompt on this page opens by asking for the Brand Brain. That is deliberate. Without it, an agent invents a brand, and you spend longer correcting the invention than you would have spent writing the piece yourself. With it, the agent argues from your positioning, your proof and your voice rules, and the correction loop collapses to a sentence or two.

If you are in a Claude Project with the Brand Brain in knowledge, write "Use the Brand Brain in project knowledge" and carry on. If you are in a bare chat, paste the filled Brand Brain directly above the agent prompt.