Documentation
Everything you need to deploy the AppXen Platform appliance in your AWS account and connect AI agents to MCP servers, knowledge and workflows through it.
AppXen Platform
Deploy and operate the appliance in your own AWS account.
Overview
What the appliance is, what its CloudFormation stack creates, what it costs
Deploy
The eight stack parameters, the Amazon Bedrock prerequisite, and your first sign-in
Connect & Use
Connect Claude Code, Claude Desktop and Cursor; API keys; register MCP servers; web search
Operate
appxenctl commands, backups, changing settings later, and upgrade notes
Network & IAM
Open ports, outbound connections, and what the instance role can do
Orchestrator Quickstart
Create your first multi-agent workflow and run it in minutes.
Gateway Quickstart
Create an API key and connect your first MCP server.
RAG Engine Quickstart
Upload your first document and run a semantic search query.
Install the CLI
Manage workflows and knowledge from the terminal.
Agent Orchestrator
Quickstart Guide
Create your first workflow, compile it, and run an execution
MCP Tool Integration
NewGive agents access to RAG search, media storage, web search, and any MCP server you register
Workflow Markdown DSL
Frontmatter, agents, steps, dependencies, fan-out loops, and approval gates
API Reference
REST API endpoints for workflow CRUD, execution, and monitoring
Model Registry
Supported models, logical names, and the Amazon Bedrock prerequisite
Console Guide
Using the dashboard editor, workflow gallery, and execution monitoring
Pricing
From $0.149 per instance-hour or $949 per year, plus your own AWS and Amazon Bedrock usage
MCP Gateway Pro
Quickstart Guide
Deploy the appliance, create an API key, and connect your first server
Connecting MCP Servers
Add remote MCP servers like GitHub, Neon, Supabase, and more
REST API Proxy
Turn any REST API with an OpenAPI spec into MCP tools your agents can call
API Reference
MCP protocol endpoints, authentication, and response formats
Media & Storage
Upload documents, images, and videos for your AI agents to access
Extensions & Built-in Services
NewEnable the CloudWatch extension, and use the built-in RAG, Media Storage and web search services
Using with Claude Desktop
Configure Claude Desktop to use your MCP Gateway
Using with OpenAI Codex
Configure OpenAI Codex to use your MCP Gateway over Streamable HTTP
Pricing
Included in the appliance; what the AWS Marketplace price covers and what your own AWS usage adds
RAG Engine
Getting Started
Upload your first document and run a semantic search query
How It Works
The ingestion pipeline: upload, parse, chunk, embed, index, search
Supported Formats
PDF, DOCX, HTML, Markdown, CSV, JSON, images, and more
Pricing
Included in the appliance; your own S3, DynamoDB and Amazon Bedrock usage is billed by AWS
API Reference
REST API endpoints and MCP tool definitions
Concepts
Need Help?
Can't find what you're looking for? We're here to help.