This guide will help you understand the core building blocks that make MeshAgent a powerful platform for building, deploying, and operating AI agents — collaboratively, securely, and at any scale. By the end of this document you will understand how Agents, Tools, and Rooms work together to create seamless AI powered experiences.

​
Key Concepts at a Glance: Agents, Tools, and Rooms

MeshAgent enables people, AI agents, and external capabilities (tools) to meet in a secure Room where they can collaborate in real-time, share context, and work together toward common goals.

  • Agents: Language model (LLM) powered entities that can understand natural language, make decisions, and execute goals over multiple iterations. Agents are typically given a persona (aka system prompt or rules) and access to tools to help them understand and perform their job.
  • Tools: Tools extend a model’s capabilities by allowing the model to take action on your behalf — something an LLM itself cannot do. Tools can be registered in the Room and called by human and agent participants. Agents can be equipped with specific tools, and human participants can invoke registered tools.
  • Rooms: A temporary, secure collaboration space that automatically spins up when the first participant arrives and vanishes when the last one leaves. All context, conversations, and shared resources live within the Room. Rooms are provisioned on demand and can scale up or down as your demand changes.

​
Agents: Your AI Teammates

​
What are Agents?

Agents are backed by an LLM plus a reasoning loop which allows the agent to continue working until a task is complete.

Agents have rules or system prompts which give them insight into their role (e.g. “You are a Document Writer for MeshAgent, an AI startup that provides the missing infrastructure to scale AI Agents”) and are equipped with tools to help them complete their task (e.g. DocumentCreation tool, WebSearch tool, SendEmail tool, etc.).

​
Types of Agents in MeshAgent

  • Chat Agents: Handle text-based conversations and queries
  • Voice Agents: Process spoken language and respond with voice
  • Task Runners: Agents with structured inputs and outputs. A Task Runner might be given to a Chat or Voice Agent as a tool. Task Runners are helpful when you need a certain format to be followed.

Alternatively, you can run agents in your agent framework of choice inside of a MeshAgent Room.

​
Tools: Extending Agent Capabilities

​
What are Tools?

Tools are specialized functions that extend what agents can do beyond basic conversation. MeshAgent ships prebuilt, reusable toolkits for common Room interactions (e.g., file and document handling, media generation, dynamic UI interactions) and high-demand tasks (e.g., web search or document-to-Markdown conversion).

MeshAgent toolkits group together related tools and pass the set of tools to the agent for use.

​
Built-in toolkits

  • Document Authoring: Create and edit documents in the room
  • UI Interactions: Interact with the user through the UI (create and generate content for forms to gather additional information to help the agent execute the request)
  • Media Generation: Create images, videos, and presentations
  • Document Processing: Convert documents from PDF to Markdown

​
Custom tools and toolkits

  • Model Context Protocol (MCP) Support: Integrate any MCP compatible service into a Room. Simply wrap the MCP image and launch it as a service inside a MeshAgent Room.
  • Fully Custom Tools: You can write your own custom tools and toolkits with MeshAgent. Tools require an __init__ and execute function and should return a specific data type (e.g. TextResponse, JsonResponse).

​
How Tools Work

  • Toolkits are passed to an Agent and the Agent determines when they need to invoke a tool to perform a task.
  • When an agent invokes a tool, the tool logic is executed and the results are streamed back to the agent.
  • Tools are centrally managed like agents, allowing you to deploy them once and reuse them effortlessly within a project.

​
Rooms: The Secure, Scalable Collaboration Hub

​
What are Rooms?

MeshAgent Rooms provide a shared workspace for agents and participants to collaborate in real time. Each room includes a participant list, presence tracking, and a mechanism for sharing information. Because MeshAgent is designed for seamless integration, rooms are provisioned on demand as participants connect. This eliminates the complexity of syncing rules and allows agents and people to quickly spin up a shared environment to collaborate in.

  • Temporary by design
    A room is provisioned when the first participant joins and deprovisioned when the last participant leaves. It only exists to facilitate communication during the session.

  • No state persistence
    The MeshAgent admin console logs session activity for debugging, such as which agents participated and what questions were asked. However, the state of the room will be discarded at the end of the session, so any data that should persist after the session ends needs to be written to an external storage provider.

  • Document management
    A room can help participants discover, sync, and share documents when they connect to it, but the room only temporarily stores MeshDocuments in memory while they are open. When documents are modified, or a session ends, the changes are written to an external storage provider. By default, MeshAgent Cloud provides a storage folder (named for the room) where documents are kept. For more complex scenarios, you can configure an existing storage solution using a custom provider and save the documents in your own database or storage buckets.