Guides¶
Practical, step-by-step guides for working with MPL. Whether you are validating your first payload, connecting MPL to an existing pipeline, or running it in production, these guides walk you through real-world scenarios.
Guide Categories¶
graph LR
Guides[Guides] --> Tutorials[Tutorials<br/>Hands-on walkthroughs]
Guides --> Integration[Integration<br/>Connecting MPL to systems]
Guides --> Operations[Operations<br/>Running in production]
style Tutorials fill:#c8e6c9
style Integration fill:#fff9c4
style Operations fill:#ffe0b2
Tutorials¶
Hands-on walkthroughs that teach MPL concepts through building. Each tutorial is self-contained and takes 15-30 minutes to complete.
| Tutorial | Description | STypes Used | Difficulty |
|---|---|---|---|
| Calendar Workflow | Create and validate calendar events through the proxy | org.calendar.Event.v1 |
Beginner |
| RAG with QoM | Validate RAG queries and evaluate groundedness | eval.rag.RAGQuery.v1, eval.rag.SearchResult.v1 |
Intermediate |
| Multi-Agent Workflow | Typed communication between planner and executor agents | org.agent.TaskPlan.v1, org.agent.ToolInvocation.v1 |
Advanced |
| Creating a Custom SType | Design, register, and validate your own semantic type | Custom (org.support.Ticket.v1) |
Intermediate |
Integration¶
Guides for connecting MPL to existing systems, frameworks, and protocols.
| Guide | Description | Prerequisites |
|---|---|---|
| MCP Integration | Connect MPL as a governance layer for MCP servers | Running MCP server |
| A2A Integration | Add semantic governance to Agent-to-Agent protocols | A2A-compatible agents |
| LangChain Integration | Use MPL validation in LangChain tool chains | LangChain installed |
| CI/CD Pipeline | Validate schemas and assertions in your build pipeline | GitHub Actions or similar |
Coming Soon
Integration guides are under active development. Check back for updates or contribute on GitHub.
Operations¶
Guides for deploying, monitoring, and maintaining MPL in production environments.
| Guide | Description | Audience |
|---|---|---|
| Production Deployment | Configure the proxy for high-availability production use | Platform Engineers |
| Monitoring and Alerting | Set up Prometheus, Grafana, and alerting for QoM metrics | SRE / DevOps |
| Schema Governance | Manage SType lifecycle across teams | Architects |
| Troubleshooting | Diagnose common issues with validation, QoM, and the proxy | All |
Coming Soon
Operations guides are under active development. Check back for updates or contribute on GitHub.
Prerequisites¶
Before starting any guide, ensure you have:
- MPL CLI installed (Installation Guide)
- MPL Proxy running on port
9443(Quick Start) - Python 3.10+ or Node.js 18+ for SDK examples
- Docker (optional, for containerized setups)
Recommended Path¶
Suggested Order
- Start with the Calendar Workflow tutorial to learn the basics
- Move to RAG with QoM to understand quality metrics
- Try Multi-Agent Workflow for complex orchestration
- Create your own type with Custom SType
- Then explore Integration and Operations guides for production deployment
Conventions Used in Guides¶
Throughout these guides, you will see:
- Tabbed code blocks with Python, TypeScript, and curl examples
- Admonitions highlighting tips, warnings, and important notes
- Mermaid diagrams showing data flow and architecture
- Full request/response examples you can copy and run directly
All examples assume the MPL proxy is running at http://localhost:9443 and metrics are available at http://localhost:9100/metrics.