Skip to main content
INFORMATIVEDRAFTDocumentation Governance
Truth Source: Repository schemas and tests are authoritative.

External Governance & Trust Overview

What MPLP Is NOT​

Before evaluating MPLP, it is important to understand what the protocol explicitly does not claim to be:

MPLP Is NOTExplanation
A ProductMPLP is a protocol specification, not a software product
A CompanyMPLP is governed by an open committee (MPGC), not a corporation
A Certification BodyMPLP does not certify vendors or issue compliance badges
A Vendor Lock-inMPLP is vendor-neutral by design
A RuntimeMPLP defines semantics, not implementations

What MPLP IS​

MPLP ISExplanation
An Open ProtocolFreely available specification for AI agent lifecycle management
Vendor-NeutralAny vendor can implement without approval
GovernedChanges require formal RFC and MPGC voting
VersionedSemantic versioning with frozen guarantees
ObservableAll agent actions are traceable by design

Trust Model​

Self-Declaration Model​

MPLP uses a self-declaration model for conformance:

  1. Implementers declare their own conformance level
  2. MPLP provides validation tools (schema validators, invariant checkers)
  3. Users evaluate implementations using MPLP's published criteria
  4. MPGC does NOT audit, certify, or endorse implementations

No "MPLP Certified" Claims​

Governance Transparency​

ArtifactLocationPurpose
Protocol SpecificationThis documentation siteDefinitions
Reference ImplementationGitHub repositorySDK and runtime examples
Governance Policies/12-governance/Change processes
RFC Archive/mips/Proposal history

How to Evaluate MPLP​

For detailed evaluation criteria, see: