MorningAI agents are a marketing team that already knows your brand. One agent handles the whole conversation: it can read your Gmail, Google Drive, and Google Calendar; query your Shopify store, Google Analytics, and Search Console; search and update your product catalog; talk to your customer personas; research the live web; render charts; generate images; and produce downloadable documents — all in the same turn, with your permission and under your control.
This article is the map of everything agents can do today, with links to a focused guide for each capability.
One Agent, Many Specialist Skills
You don't pick tools or build workflows. You describe the outcome, and the agent decides which of its capabilities to use — and can chain several in a single message. Ask "Pull my store's sales for last month, chart revenue by week, and build me a 6-slide deck for Monday's team meeting" and you'll watch the progress chip walk through each step: Reading your store → Building the chart → Generating file. When the turn finishes, the chip reads "Thought for Xs" and expands to show every step the agent took. You can stop a turn at any time.
What Agents Work With
| What | What it gives the agent |
|---|---|
| Your DNA | BrandDNA voice and facts applied to every conversation automatically; @-mention any product or persona to attach it to a turn |
| Your systems | Six connectors: Gmail, Google Drive, and Google Calendar (personal), plus Shopify, Google Analytics, and Search Console (shared by your brand team) |
| Specialist skills | Documents in nine formats, interactive charts, image generation, live web research, and product catalog work |
| Your controls | Per-tool permissions (Always allow / Needs approval / Blocked) and in-chat approval cards before any write runs |
| Where work lands | Your file library, the in-chat preview panel, the File Editor, and Home — one click from published |
The Composer Basics
The same composer lives on both Home and Chat, and its hint tells you most of what you need: "/ for playbooks · @ for products & personas".
- Type @ to attach a specific product or customer persona to your message. The agent already knows your catalog — mentioning makes a turn about that product or that audience. See What Agents Know.
- Type / to run a playbook — a reusable instruction set you or a teammate saved. See Playbooks.
- Drop files into the composer and the agent reads them — briefs, spreadsheets, reference decks — and can come back to them later in the conversation.
- Pick a voice. The welcome screen has two tabs: Marketing Generalist (with seven expert modes like SEO, Social Media, and Analytics) and Customer Insights, where the agent answers as one of your customer personas.
Your Permission, Your Control
The default posture is simple: reads are allowed, every connector write pauses for your approval, and you can tighten or loosen any individual tool at any time. Some limits are hard guarantees, not settings — the agent can draft emails but can never send one, and it can never delete anything: not emails, not files, not calendar events, not products. Trust is a feature; read the full permissions and approvals guide.
Where the Work Lands
Agent output isn't a wall of text you reformat elsewhere. Generated documents land in your file library, preview inline in the conversation, and — for PDFs and decks — open in the File Editor for finishing touches. On Home, the Recent Work section has a By agents tab that collects agent-produced work, with the publish composer opening in place — one click from agent output to a scheduled post.
Explore What Agents Can Do
- Agent Use-Case Gallery — ten copy-pasteable recipes, from inbox triage to a store-performance deck. Start here if you want ideas.
- Generate documents in nine formats, then edit PDFs and decks in the File Editor.
- Ask your store and analytics anything and get answers with interactive charts rendered in the conversation.
- Connect Gmail, Drive, and Calendar, Search Console and Google Analytics, or your Shopify store.
- Watch live web research as it happens, or test messaging on your customer personas before it ships.
- Encode your team's process as playbooks, and understand when to use Studio vs Agents.
A Quick Word on Brainstorming
The simplest way to start is still a conversation. Describe a goal — "I want to promote a new product to Gen Z snackers" or "We need a fun back-to-school campaign for our beverage line" — and iterate on the ideas that come back. The agent already knows your brand voice, your catalog, and your personas, so you can skip the context-pasting and go straight to the thinking. @-mention a persona to aim the ideas at a specific audience.
Frequently Asked Questions
Do agents run on a schedule? No. Agents work on demand, in conversation. For social content that runs autonomously on a cadence, see Done For You — a separate engine with its own setup.
Can the agent send emails or publish posts? No. Email is draft-only by design — you always send from Gmail yourself. Publishing happens through the composer and calendar; agent work reaches it via Home → Recent Work.
Do I need to set anything up? Your DNA grounding is built during brand setup, so agents are useful immediately. Connectors are optional and each takes about a minute to connect.
Next Steps
Open a conversation and ask for something with a finished shape — a deck, a spreadsheet, a plan. Then browse the use-case gallery for nine more ideas your team can run this week.
