Skip to main content
PraisonAI has three distinct MCP roles across three packages — connecting to servers, serving a small tool set, and hosting the full capability registry.
The snippet above is the client layer. Serving your own agents uses the light or heavy layer instead.

Which Layer Do I Need?

The Layers

Each layer maps to one package and one job.

Client

Connect an agent to any external MCP server and use its tools.

Light server

Serve a small, config-driven tool set from the terminal CLI.

Heavy host

Serve the full PraisonAI capability registry, recipes, and auth.

Best Practices

Use the agents-tier MCP class to consume external tools. Use praisonai-mcp to expose your own agents to MCP clients.
praisonai serve mcp is the lightest way to publish a handful of config-driven tools without the heavy host.
Reach for praisonai-mcp when you need OAuth 2.1 / OIDC, HTTP-stream transport, or recipe serving.

praisonai-mcp Package

Standalone heavy-host package guide.

MCP Integration

Connect agents to external MCP servers.

PraisonAI MCP Server

Heavy MCP host reference.

Package Tiers

How the seven packages stack.