Follow these steps to run the MCP server agent and chatbot toolkit inside a cloud Room for quick testing. When the Room closes (e.g., is inactive), the tools and bots are also automatically removed. Required Environment Variables:
  • AWS_ACCESS_KEY_ID=YOUR_ACCESS_KEY_HERE
  • AWS_SECRET_ACCESS_KEY=YOUR_SECRET_ACCESS_KEY_HERE

​
Steps

  1. Install Meshagent CLI 
    pip install "meshagent[all]"
  2. Sign Up and Authenticate
  3. Start the MCP Server Agent in the Room 
    meshagent service test --room=test --role=agent \
  --image=meshagent/mcp-aws-kb-retrieval-server:latest \
  --env MESHAGENT_PORT=8001 \
  --env AWS_ACCESS_KEY_ID=YOUR_ACCESS_KEY_HERE \
  --env AWS_SECRET_ACCESS_KEY=YOUR_SECRET_ACCESS_KEY_HERE \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-aws-kb-retrieval-server-service-test
    • This command launches a temporary Meshagent Room with the MCP server available as an agent. The Room closes if it goes inactive.
  4. Start the Chatbot with the Tool in the Same Room 
    meshagent chatbot join --room=test \
  --agent-name=mcp-aws-kb-retrieval-server \
  --name=mcp-aws-kb-retrieval-server \
  --toolkit=mcp-aws-kb-retrieval-server
    • This command launches a chatbot in the Room, making the aws-kb-retrieval-server tools available to participants.
    • Multiple toolkits can be enabled in the same Room.
  5. Visit the Room to Interact
    • The output from the previous command will provide a Room link. Open it in your browser, and send messages to the chatbot agent to interact with the MCP server tools.

​
Project-Level Deployment (Persistent, for Production)

To have the MCP server and chatbot automatically available in all new Rooms created in a Meshagent project (no need to run/attach them each time), set up project-level services.
  1. Create Persisting MCP Server Agent: 
    meshagent service create --role=agent \
  --image=meshagent/mcp-aws-kb-retrieval-server:latest \
  --env MESHAGENT_PORT=8001 \
  --env AWS_ACCESS_KEY_ID=YOUR_ACCESS_KEY_HERE \
  --env AWS_SECRET_ACCESS_KEY=YOUR_SECRET_ACCESS_KEY_HERE \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-aws-kb-retrieval-server-service
  2. Create Persisting Chatbot Service: 
    meshagent service create \
  --image="meshagent/cli:latest" \
  --port="num=9001 path=/agent liveness=/ type=meshagent.callable participant_name=mcp-aws-kb-retrieval-server-chatbot" \
  --env="MESHAGENT_PORT=9001" \
  --name="mcp-aws-kb-retrieval-server-chatbot-service" \
  --command="meshagent chatbot service --agent-name=mcp-aws-kb-retrieval-server-chatbot --toolkit=mcp-aws-kb-retrieval-server"
    • Whenever a Room is created in your project, the MCP server and chatbot/toolkit will auto-join, providing seamless capability for all project Rooms.

​
Tools Available

​
Tools Provided by this Server

ToolShort Description
retrieve_from_aws_kbPerforms retrieval from the AWS Knowledge Base using the provided query and Knowledge Base ID.

​
Tool Details

​
Tool: retrieve_from_aws_kb

Performs retrieval from the AWS Knowledge Base using the provided query and Knowledge Base ID.
ParameterTypeDescription
knowledgeBaseIdstringThe ID of the AWS Knowledge Base
querystringThe query to perform retrieval on
nnumberoptionalNumber of results to retrieve

​
Meshagent & Resources