INFORMATIVESTABLE
Truth Source: Repository schemas and tests are authoritative.
Introduction to MPLP
Welcome to the Multi-Agent Lifecycle Protocol (MPLP) documentation. This section provides foundational context for understanding the protocol's purpose, scope, and design philosophy.
What is MPLP?
MPLP is a vendor-neutral, schema-first specification that defines how multi-agent systems should govern their lifecycle — from initialization through execution to completion and audit.
MPLP is not a runtime, framework, or product. It is a normative specification that enables interoperability across diverse agent implementations.
Documentation Structure
| Document | Purpose |
|---|---|
| Protocol Overview | High-level introduction to MPLP concepts |
| Docs Map | Navigation guide for the documentation |
| API Quick Reference | Schema and endpoint summary |
| Glossary | Terminology definitions |
| POSIX Analogy | Conceptual mapping to familiar OS concepts |
Key Principles
- Schema is Truth — All behavior is defined by JSON schemas, not prose
- Vendor Neutral — No dependency on specific runtimes or frameworks
- Audit-Ready — Every lifecycle event is traceable and verifiable
- Substrate Agnostic — Works across LangChain, AutoGen, CrewAI, etc.
Getting Started
- Start with the Protocol Overview for conceptual grounding
- Explore the Specification for normative requirements
- See Examples for practical implementations
Disambiguation
MPLP = Multi-Agent Lifecycle Protocol (not a license, not a product)