Easily try the MCP server and its tools in a temporary Meshagent Room in the cloud:

Prerequisites


1. Environment Variables

Set your environment variables (replace the token as needed):
  • E2B_API_KEY – Your E2B API key

2. Start the MCP Server as a Meshagent Room Service

meshagent service test \
  --room=test \
  --role=agent \
  --image=meshagent/mcp-e2b:latest \
  --env MESHAGENT_PORT=8001 \
  --env E2B_API_KEY=YOUR_API_KEY_HERE \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-e2b-service-test
  • Starts a Meshagent Room with the MCP server running as a test agent.
  • Service and chatbot clean up automatically when the Room closes or goes inactive.

3. Join with a Chatbot Agent

meshagent chatbot join \
  --room=test \
  --agent-name=mcp-e2b \
  --name=mcp-e2b \
  --toolkit=mcp-e2b
  • Launches a chatbot in the same Room, provisioned with the e2b toolkit.
  • You will receive a browser link to the Room—multiple toolkits can be used in the same Room.

4. Interact with the MCP Server Tools

  • Open the Room link in your browser.
  • Send a message to the agent and interact with the MCP server tool, e.g., ask it to run Python code!

Environment Variables (Summary)

  • E2B_API_KEYYour E2B API Key (required)
  • MESHAGENT_PORTSet to 8001 for the MCP server

🔒 Project-Level Deployment (Persistent Integration)

Set up always-available MCP server and chatbot services in all Meshagent Rooms for your project—great for production or consistent workflows (no separate hosting needed):

1. Create Persistent MCP Service

meshagent service create \
  --role=agent \
  --image=meshagent/mcp-e2b:latest \
  --env MESHAGENT_PORT=8001 \
  --env E2B_API_KEY=YOUR_API_KEY_HERE \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-e2b-service

2. Create Persistent Chatbot Service

meshagent service create \
  --image="meshagent/cli:latest" \
  --port="num=9001 path=/agent liveness=/ type=meshagent.callable participant_name=mcp-e2b-chatbot" \
  --env="MESHAGENT_PORT=9001" \
  --name="mcp-e2b-chatbot-service" \
  --command="meshagent chatbot service --agent-name=mcp-e2b-chatbot --toolkit=mcp-e2b"
  • Both services will start in every Room created in the project, allowing all users instant access to the MCP tools and chatbot.

🛠️ Tools Available

run_code

Run python code in a secure sandbox by E2B. Short Description:
Run python code in a secure sandbox by E2B.
Details:
  • Parameters:
    • code (string): Python code to be executed.
  • Syntax: Jupyter Notebook style.

MCP Server Info