PAGE_ROLE: ORIENTATION HUB

Start here.

A short orientation to Jearon Wong's research ledger.

The site follows one question: how does agent work move from human intent to an accepted outcome with responsibility and evidence still attached?
READING_ORDER: ORIENTATION_HUB

Start with one question.

Start with the category, then choose the record type you need: field, protocol, evidence, or argument.

Read the three-layer chainOpen the field definitionInspect the proof path
  1. 01Operational problem
  2. 02Three-layer chain
  3. 03System map
  4. 04Choose a path
  5. 05Boundary
SEMANTIC_RECORD: DEFINITION / BOUNDARY / EVIDENCE / CITATION
DEFINITION
AI Agent Lifecycle defines the accountable lifecycle of agent work from intent to accepted outcome.
BOUNDARY
This site presents authored field analysis, public research, protocol work, runtime and delivery projects, and evidence surfaces. It does not claim that MPLP is a complete Agent OS, that it automatically detects every Intent Drift, or that it provides certification, legal-compliance proof, insurance approval, or regulatory approval.
EVIDENCE
The lifecycle page, protocol project, research ledger, and evidence surface are the canonical records behind this orientation route.
CITATION
Use the linked canonical records when citing a definition, protocol role, research claim, or evidence artifact.
SOURCE_LEVEL
AUTHORED_ORIENTATION
UPDATED
2026-08-28
01_OPERATIONAL_PROBLEM

The operational problem

An agent can execute a task and still leave the work without a durable account of intent, authority, review, or acceptance. Execution is not Delivery.

02_THREE_LAYER_CHAIN

The missing lifecycle

AI Agent Lifecycle defines the accountable lifecycle of agent work from intent to accepted outcome. Agentic Delivery names the layer between agent execution and accountable outcomes.

  1. 01AI Agent LifecycleThe field definition: the unit of accountability.
  2. 02Agentic DeliveryThe category for carrying intent through to an accepted outcome.
  3. 03MPLPThe open, vendor-neutral lifecycle protocol path that makes the work explicit and auditable.
VISUAL_MODEL: PROJECTS

One reading order, four proof roles.

Start with the field, inspect the protocol and runtime paths, then follow delivery and evidence records.

  1. 01Protocol pathMPLP lifecycle semantics
  2. 02Runtime pathState and execution substrate
  3. 03Proof pathApplied delivery records
  4. 04Evidence surfaceAdjudication and citation
Read left to right on wide screens and top to bottom on small screens. Explanatory sequence for this page.
03_SYSTEM_MAP

How the named projects relate.

The projects are distinct proof paths around one category. The project roles do not redefine MPLP protocol authority.

01MPLPProtocol path02Cognitive OSRuntime path03SoloCrewDelivery path04Validation LabEvidence path
04_CHOOSE_A_PATH
01FieldRead the category definition first.Open record ->02ProtocolInspect how lifecycle responsibility is represented.Open record ->03EvidenceFollow the records and artifacts that make a claim inspectable.Open record ->04ArgumentRead the essays that explain why the distinction matters.Open record ->
05_BOUNDARY

Read the scope before the claims.

This site presents authored field analysis, public research, protocol work, runtime and delivery projects, and evidence surfaces. It does not claim that MPLP is a complete Agent OS, that it automatically detects every Intent Drift, or that it provides certification, legal-compliance proof, insurance approval, or regulatory approval.