The following steps show how to launch a Pulumi MCP server and a Meshagent-connected chatbot into a shared “Room” for collaborative, cloud-hosted testing.

​
1. Install Meshagent Python Package

pip install "meshagent[all]"

​
2. Sign Up and Authenticate

Create an account and link your Meshagent CLI as described here.

​
3. Start the MCP Server in a Meshagent Room

meshagent service test --room=test --role=agent \
  --image=meshagent/mcp-pulumi:latest \
  --env MESHAGENT_PORT=8001 \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-pulumi-service-test
  • This will start a Pulumi MCP server inside the test Room.
  • Rooms auto-close on inactivity, tearing down the server/tooling.

​
4. Start a Chatbot Connected to Pulumi in the Same Room

meshagent chatbot join --room=test --agent-name=mcp-pulumi --name=mcp-pulumi --toolkit=mcp-pulumi
  • This launches a chatbot with Pulumi toolkit access inside the Room.
  • Multiple toolkits may be attached as needed.
  • The CLI output provides a direct room link.

​
5. Try It Out

Visit the provided Room link in your browser. Interact with the mcp-pulumi agent via chat to access MCP server tools in real time!

​
Project Level Deployment (Persistent, Auto-Join Services)

For production use, deploy persistent MCP services and chatbots at the project level. Every new Room in the project will have these services available automatically.
  • No need to manually launch servers/agents for each Room.
  • Recommended for teams and ongoing workflows.

​
1. Create Persistent MCP Server Service

meshagent service create --role=agent \
  --image=meshagent/mcp-pulumi:latest \
  --env MESHAGENT_PORT=8001 \
  --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
  --name=mcp-pulumi-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-pulumi-chatbot" \
  --env="MESHAGENT_PORT=9001" \
  --name="mcp-pulumi-chatbot-service" \
  --command="meshagent chatbot service --agent-name=mcp-pulumi-chatbot --toolkit=mcp-pulumi"

​
Tools Available

Below are the tools provided by the Pulumi MCP Server.

​
Tool Summary

ToolShort Description
deploy-to-awsDeploy application code to AWS by generating Pulumi infrastructure.
pulumi-cli-previewRun pulumi preview for a given project and stack
pulumi-cli-refreshRun pulumi refresh for a given project and stack
pulumi-cli-stack-outputGet the output value(s) of a given stack
pulumi-cli-upRun pulumi up for a given project and stack
pulumi-registry-get-functionReturns information about a Pulumi Registry function
pulumi-registry-get-resourceReturns information about a Pulumi Registry resource
pulumi-registry-get-typeGet the JSON schema for a specific JSON schema type reference
pulumi-registry-list-functionsList all function types for a given provider and module
pulumi-registry-list-resourcesList all resource types for a given provider and module

​
Tool Details

​
deploy-to-aws

Deploy application code to AWS by generating Pulumi infrastructure. This tool automatically analyzes your application files and provisions the appropriate AWS resources (S3, Lambda, EC2, etc.) based on what it finds. No prior analysis needed - just invoke directly.

​
pulumi-cli-preview

Run pulumi preview for a given project and stack
Parameters:
  • workDir (string): The working directory of the program.
  • stackName (string, optional): The associated stack name. Defaults to ‘dev’.

​
pulumi-cli-refresh

Run pulumi refresh for a given project and stack
Parameters:
  • workDir (string): The working directory of the program.
  • stackName (string, optional): The associated stack name. Defaults to ‘dev’.

​
pulumi-cli-stack-output

Get the output value(s) of a given stack
Parameters:
  • workDir (string): The working directory of the program.
  • outputName (string, optional): The specific stack output name to retrieve.
  • stackName (string, optional): The associated stack name. Defaults to ‘dev’.

​
pulumi-cli-up

Run pulumi up for a given project and stack
Parameters:
  • workDir (string): The working directory of the program.
  • stackName (string, optional): The associated stack name. Defaults to ‘dev’.

​
pulumi-registry-get-function

Returns information about a Pulumi Registry function
Parameters:
  • function (string): The function type to query (e.g., ‘getBucket’, ‘getFunction’, ‘getInstance’)
  • provider (string): The cloud provider (e.g., ‘aws’, ‘azure’, ‘gcp’, ‘random’) or github.com/org/repo for Git-hosted components
  • module (string, optional): The module to query (e.g. ‘s3’, ‘ec2’, ‘lambda’). Optional.
  • version (string, optional): Provider version (e.g. ‘6.0.0’). Optional.

​
pulumi-registry-get-resource

Returns information about a Pulumi Registry resource
Parameters:
  • provider (string): The cloud provider (e.g., ‘aws’, ‘azure’, ‘gcp’, ‘random’) or github.com/org/repo for Git-hosted components
  • resource (string): The resource type to query (e.g., ‘Bucket’, ‘Function’, ‘Instance’)
  • module (string, optional): The module to query (e.g. ‘s3’, ‘ec2’). Optional.
  • version (string, optional): Provider version. Optional.

​
pulumi-registry-get-type

Get the JSON schema for a specific JSON schema type reference
Parameters:
  • name (string): The name of the type (e.g., ‘BucketGrant’, ‘FunctionEnvironment’)
  • provider (string): The cloud provider
  • module (string, optional): Module to query. Defaults to ‘index’.
  • version (string, optional): Provider version. Optional.

​
pulumi-registry-list-functions

List all function types for a given provider and module
Parameters:
  • provider (string): The cloud provider
  • module (string, optional): Optional module filter
  • version (string, optional): Provider version. Optional.

​
pulumi-registry-list-resources

List all resource types for a given provider and module
Parameters:
  • provider (string): The cloud provider
  • module (string, optional): Optional module filter
  • version (string, optional): Provider version. Optional.

​
Meshagent, MCP Server, and References