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

# Agent Scope

> Control which Logic agents are accessible through your MCP server

## Understanding Agent Scope

When you create an MCP server, you choose which Logic agents it can access. Logic provides two scope options: **All Agents** and **Limited**. This determines which agents AI platforms can execute through the server.

<Note>
  Scope controls *which* agents a server reaches. A separate **permission**
  setting controls *what* a client can do with them (run only, or also create
  and edit). See [Built-in Tools](/mcp/built-in-tools) for permissions.
</Note>

## Scope Types

### All Agents Scope

Access to **all agents** in your organization, including future agents. New agents are automatically available without configuration.

Best for internal tools where you trust all content to be accessible.

### Limited Scope

Access to **only agents you explicitly assign**. You control exactly which agents are exposed.

Best for external sharing, testing environments, or when you need fine-grained control. Note that new agents must be manually added.

## Next Steps

* [Create your first MCP server](/mcp/quick-start)
* [Learn about async task execution](/mcp/async-tasks)
* [Connect to an AI platform](/mcp/quick-start#step-4%3A-connect-to-an-ai-platform)