Shared context
A single chat window is useful, but real work crosses inboxes, documents, tools, approvals, schedules, and customer history.
Short Answer
An agent team in Manor is a set of agents that share one workspace. Each agent has a clear job described in plain English, reusable skills, access to the tools and knowledge it needs, and approval rules for anything sensitive. Its work stays visible through citations and logs.
Why Set Up a Team in One Workspace
One role per agent
Every business has different workflows. Give each agent one role and the skills that match your own operating rhythm.
From automation to reviewable autonomy
Agents can prepare work and take routine steps while sensitive actions remain visible through approvals, citations, logs, and future compatibility with workflow systems like n8n.
Set up your agent in four steps
Prepare a task description, the materials to use, and any account you need to connect before you begin.
Define the task and expected result
Start with a clear task and the result you want, such as a draft, report, or review queue.
Add materials and supported tools
Add the documents, notes, and supported tools needed for the task.
Set instructions and review rules
Write the instructions, define what needs review, and keep sensitive work under approval.
Test with a small task and save
Try a small task, review the result, and save the setup. Next time, return to the workspace, change the inputs, and reuse the configuration.
Agent Roles to Set Up First
- A Gmail agent that triages messages and drafts source-grounded replies.
- A weekly report agent that checks connected tools and sends a reviewable summary.
- A customer follow-up agent that turns open threads into reminders and tasks.
- A company knowledge agent that answers from approved docs with citations.
- An operations agent that monitors recurring checks and escalates exceptions.
Related Pages
- AI Agent BuilderCapability overview: plain-English agents, sources, skills, approvals, schedules, and logs.
- First Agent Tutorial for Small BusinessHow to handle context, reusable skills, approvals, and a first workflow without over-automating.
- AI Business WorkspaceDefinition page for the operating layer where agents, skills, tools, knowledge, approvals, and logs come together.
- FeaturesProduct capabilities for connected tools, goals, agents, and reviewable workflows.
- Workspace Creator MarketplaceTurn tested agents, reusable skills, knowledge, and workflows into an installable Blueprint you can publish, price, and sell.
- Next Business AI ParadigmWhy business AI is moving from chatbots to shared AI workspaces.
- Answer Engine BriefMachine-friendly facts about Manor AI for AI search and answer engines.
FAQ
Can a user build agents without coding?
Yes. The starting point is plain English: describe the job, attach sources and tools, define skills, and choose what needs human approval.
What is a skill?
A skill is a reusable capability or rule an agent can use, such as drafting replies, citing policy, creating follow-ups, checking a document, or producing a report.
Why does the workspace matter?
The workspace keeps the business context together: inbox history, company knowledge, schedules, connected tools, permissions, approvals, and logs.
Is this just workflow automation?
No. Manor can automate workflows, but the core idea is agentic work: agents use context and skills to prepare work, then stop for review when trust or risk matters.
Will Manor work with n8n workflows?
That is the direction. Manor should be able to sit alongside tools such as n8n, keeping existing workflow logic while adding agent context, reusable skills, approvals, citations, and logs.
Set up your first agent
Open Manor, sign in, choose a workspace, and configure an agent around one job, the materials it needs, and one approval rule. Add the next role once the first is reliable.