767 Readme Containers D7d88296

Next Gen UI MCP Server Container

This module is part of the Next Gen UI Agent project.

Next Gen UI Agent MCP Server container image.

Provides

• container image to easily run Next Gen UI Agent MCP server

Installation

podman pull quay.io/next-gen-ui/mcp

Usage

Locally

podman run --rm -it -p 5100:5100 --env MCP_PORT="5100" \
       --env NGUI_MODEL="llama3.2" --env NGUI_PROVIDER_API_BASE_URL=http://host.containers.internal:11434 --env NGUI_PROVIDER_API_KEY="ollama" \
       quay.io/next-gen-ui/mcp

Openshift

oc login ...
oc project next-gen-ui # or another
oc apply -f deployment.yaml

Configuration

The MCP server container can be configured via environment variables. For available env variables and their meaning see MCP Server Guide.

Dependencien necessary for openai inference provider are installed in the image.

json and rhds renderers are installed. Create child image to install additional ones.

Default values are changed for some configurations in the image!

Usage Examples

Basic Usage with Ollama (Local LLM)

podman run --rm -it -p 5000:5000 \
    --env MCP_PORT="5000" \
    --env NGUI_PROVIDER="openai" \
    --env NGUI_MODEL="llama3.2" \
    --env NGUI_PROVIDER_API_BASE_URL="http://host.containers.internal:11434/v1" \
    --env NGUI_PROVIDER_API_KEY="ollama" \
    quay.io/next-gen-ui/mcp

OpenAI API Configuration

podman run --rm -it -p 5000:5000 \
    --env NGUI_PROVIDER="openai" \
    --env NGUI_MODEL="gpt-4o" \
    --env NGUI_PROVIDER_API_KEY="your-openai-api-key" \
    quay.io/next-gen-ui/mcp

Remote LlamaStack Server

podman run --rm -it -p 5000:5000 \
    --env NGUI_PROVIDER="openai" \
    --env NGUI_MODEL="llama3.2-3b" \
    --env NGUI_PROVIDER_API_BASE_URL="http://host.containers.internal:5001/v1" \
    quay.io/next-gen-ui/mcp

MCP Sampling with Model Preferences

podman run --rm -it -p 5000:5000 \
    --env MCP_PORT="5000" \
    --env NGUI_PROVIDER="mcp" \
    --env NGUI_SAMPLING_HINTS="claude-3-sonnet,claude" \
    --env NGUI_SAMPLING_COST_PRIORITY="0.3" \
    --env NGUI_SAMPLING_SPEED_PRIORITY="0.8" \
    --env NGUI_SAMPLING_INTELLIGENCE_PRIORITY="0.5" \
    quay.io/next-gen-ui/mcp

Configuration Using Environment File

Create a .env file:

# .env file
MCP_PORT=5000
MCP_HOST=0.0.0.0
MCP_TRANSPORT=streamable-http
MCP_STRUCTURED_OUTPUT_ENABLED="false"
NGUI_COMPONENT_SYSTEM=json
NGUI_PROVIDER=openai
NGUI_MODEL=gpt-4o
NGUI_PROVIDER_API_KEY=your-api-key-here

# Example values if MCP sampling is used
# NGUI_SAMPLING_HINTS=claude-3-sonnet,claude
# NGUI_SAMPLING_COST_PRIORITY=0.3
# NGUI_SAMPLING_SPEED_PRIORITY=0.8
# NGUI_SAMPLING_INTELLIGENCE_PRIORITY=0.5

Run with environment file:

podman run --rm -it -p 5000:5000 --env-file .env quay.io/next-gen-ui/mcp

Network Configuration

For local development connecting to services running on the host machine:

• Use host.containers.internal to access host services (works with Podman and Docker Desktop)
• For Linux with Podman, you may need to use host.docker.internal or the host's IP address
• Ensure the target services (like Ollama) are accessible from containers

Links

• Documentation
• Source Codes
• Contributing