> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mindset.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent Management Studio (AMS): Quick Start Walkthrough

> Walkthrough of the Agent Management Studio

## What is AMS?

The [**Agent Management Studio (AMS)**](https://docs.mindset.ai/amp/introduction) is a visual, no-code admin portal for building, testing, and monitoring AI agents. You can use AMS to control workflow access, manage MCP servers, upload data and guardrails, shape personality and behavior, and deploy agents quickly without writing code.

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS1.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=c73eb5a4bf0cc69af4b85d26430d514b" alt="" width="2931" height="1662" data-path="images/AMS1.png" />

## AMS Key Components at a Glance

### Agents

Autonomous units tailored to a role or objective. Agents use assigned Workflows, MCP Servers, and Knowledge Contexts to act like subject matter experts.

[Learn more about agents here](https://docs.mindset.ai/amp/agents)

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS2.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=5157c57bca0b6cb3e358254d8ff38077" alt="" width="2929" height="1655" data-path="images/AMS2.png" />

When creating an agent, you need to configure all attributes.\
[More information can be found here](https://docs.mindset.ai/amp/agents)

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS311.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=d2d5fc6422b2df3130495430be50f86e" alt="" width="2914" height="1655" data-path="images/AMS311.png" />

### Knowledge Contexts

Bundles of files, links, and data that an agent can retrieve at runtime. This is the agent’s dynamic reference library.

[Find more information about Knowledge Contexts here](https://docs.mindset.ai/amp/contexts)

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS4.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=4c3a4b028bbed5cf7e65de9a8ba891e7" alt="" width="2930" height="1665" data-path="images/AMS4.png" />

### MCP Servers

Model Context Protocol (MCP) servers standardize how agents talk to tools, APIs, and data sources. They bundle tools and expose Workflows to agents in a consistent way.

[Learn more about MCP Servers here](https://docs.mindset.ai/amp/MCPservers)

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS58.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=337613bb1a9e8f945b32e2fdceceec88" alt="" width="2930" height="1645" data-path="images/AMS58.png" />

### Workflows

Workflows transform how you build agent capabilities. Instead of complex development, you describe what you want your agent to do in natural language and the system generates functional tools automatically.

[More about Workflows here](https://docs.mindset.ai/amp/tools)

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS6.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=5438f3a5d0a8273b8c8435cf9c7d6b87" alt="" width="2932" height="1667" data-path="images/AMS6.png" />

***

## End-to-End: Build and Deploy an Agent

### Step 1: Create a Workflow (Workflow Builder)

* Go to **Workflows → New Workflow**
* Describe what the workflow should do in plain language
* Refine until the preview looks correct

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS7.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=e271e61fa9aff7ec74923bf72451dcfc" alt="" width="2928" height="1640" data-path="images/AMS7.png" />

### Step 2: Add the Workflow to an MCP Server

* Open **MCP Servers** and create or select a server

  <img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS58.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=337613bb1a9e8f945b32e2fdceceec88" alt="" width="2930" height="1645" data-path="images/AMS58.png" />
* In **Workflows**, attach the Workflow to the MCP server

  <img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS9.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=3101af4bdf001ba5ea449f2a2a2620e4" alt="" width="2927" height="1657" data-path="images/AMS9.png" />

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS10.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=477396551a3782c06f1bd65280b9a372" alt="" width="2928" height="1660" data-path="images/AMS10.png" />

* Save to expose the workflow via this MCP server

### Step 3: Create your Agent

* Go to **Agents → New Agent**
* Set name, role, or objective, and personality or guardrails
* Save

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS311.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=d2d5fc6422b2df3130495430be50f86e" alt="" width="2914" height="1655" data-path="images/AMS311.png" />

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS12.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=4425250c663c891dac1b246a6438ade2" alt="" width="2931" height="1657" data-path="images/AMS12.png" />

### Step 4: Create a Knowledge Context

* Go to **Knowledge Contexts → New Context**
* Name and describe your new context

  <img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS13.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=1dce2d28da1e6d2843b76ac906dc037a" alt="" width="2929" height="1656" data-path="images/AMS13.png" />
* Upload content into the knowledge context

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS14.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=6fc57dfe12202eeef551bce8f1463d97" alt="" width="2929" height="1667" data-path="images/AMS14.png" />

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS15.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=206e521409d042e39a945ece5471bbfe" alt="" width="2924" height="1662" data-path="images/AMS15.png" />

### Step 5: Attach the Knowledge Context to the Agent

* Open your **Agent → Contexts tab**
* Add one or more Knowledge Contexts

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS16.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=63eb4078ce49201562422e353bfea666" alt="" width="2928" height="1666" data-path="images/AMS16.png" />

### Step 6: Attach the MCP Server to the Agent

* Go to **Agent → MCP Servers tab**
* Add one or more MCP servers to the agent

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS17.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=31496522c8c931ef252881239f8ce03a" alt="" width="2931" height="1667" data-path="images/AMS17.png" />

### Step 7: Deploy the Agent

* Go back to the **dashboard** to see your live agents
* Select the agent you want to deploy
* Test it out if needed

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS18.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=6d1beab28cc9ec360f5c6376e712dfca" alt="" width="2928" height="1659" data-path="images/AMS18.png" />

Then, go to the **Deployment tab** to complete deployment.

<img src="https://mintcdn.com/mindsetai/ZhHgvQv0tI5DT0_U/images/AMS19.png?fit=max&auto=format&n=ZhHgvQv0tI5DT0_U&q=85&s=d26e65ca9f20b262fe72b9328fc12ae6" alt="" width="2932" height="1661" data-path="images/AMS19.png" />

[Learn more about deploying your agents here](https://docs.mindset.ai/amp/Deployment)

## Notes

To view agents on the demo page:

* Anonymous access must be enabled
* The agent must be set to Live