AGCP Runtime Governance Training
Learn the core ideas of deterministic runtime governance through six foundational readings, companion video briefings, and short knowledge checks.
This self-paced series is designed for professionals, students, architects, engineers, governance practitioners, risk teams, auditors, and researchers who want to understand how governance moves from policy and oversight into execution-time control for AI-enabled, agentic, autonomous, multi-agent, and programmatic systems.
Begin with two short orientation resources: Runtime Governance in Plain English, followed by the AI Runtime Governance Overview. Then work through the six modules in sequence. For each module, watch the briefing, read the source paper, and complete the corresponding Check Your Knowledge quiz. After the six readings, use the AGCP Runtime Governance Assessment Readiness Check and then return to the AI Runtime Governance Learning Guide for an integrated review of the complete reading series.
How to Use This Self-Paced Series
Video Briefing
Begin each module with the embedded video for an orientation to the problem, architecture, core concepts, and major takeaways.
Core Reading
Read the source paper in full. The papers are the authoritative basis for the concepts covered in the learning series.
Check Your Knowledge
Complete the short quiz after each reading to confirm understanding of the paper’s core distinctions, architecture, and implications.
This is a learning pathway, not an AGCP conformance assessment. The quizzes are educational knowledge checks. The readiness check at the end is a pre-assessment diagnostic and does not establish an AGCP conformance result.
Runtime Governance Orientation
Before beginning the six core papers, start with these two orientation resources. They provide a plain-English introduction to the runtime-governance problem, followed by a more structured overview of the architecture and concepts developed across the reading series.
Runtime Governance in Plain English
Begin here for a concise, accessible explanation of why AI runtime governance is needed and what it means to govern consequential AI-enabled actions at execution time.
AI Runtime Governance Overview
Continue with this overview before starting the papers. It provides a broader architectural orientation to runtime governance and prepares the learner for the concepts developed in the six core readings.
These two orientation resources do not have associated knowledge-check quizzes. The quizzes begin with Module 1 and correspond to the six core papers.
AI Runtime Governance Learning Guide
Use the AI Runtime Governance Learning Guide as a companion to the six-paper series. The guide explains the major lessons in each paper, highlights the concepts to watch for, and shows how the readings build on one another as a coherent runtime-governance architecture.
You can review the guide before beginning the series for orientation, revisit it between modules to reinforce the main lessons, or use it after completing the readings as a structured synthesis of the full learning pathway.
Six Core Readings
The series moves from the broad execution-governance problem, through runtime architecture and the AGCP control-plane model, into formal execution semantics, operating semantics, and multi-agent enterprise application.
Runtime Execution Governance for AI Systems
A Cross-Platform Synthesis and Architectural Framework
Willis, J. M. (2026). Runtime Execution Governance for AI Systems: A Cross-Platform Synthesis and Architectural Framework. Sustainable Future Tech, Inc. https://doi.org/10.5281/zenodo.19341177
Focus: Why AI governance must address the point where a proposed consequential action becomes operational effect.
- Proposed consequential action and execution governance
- Admissibility, context, and canonical state
- Commit boundary and governance receipt
- Action-level runtime control rather than model-level review alone
Companion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Test your understanding of the paper’s central architecture and distinctions.
Take the QuizRuntime Governance Architecture
Consistent Governance Execution for Enterprise Systems and Autonomous Agents
Willis, J. M. (2026). Runtime Governance Architecture: Consistent Governance Execution for Enterprise Systems and Autonomous Agents. Sustainable Future Tech, Inc. https://doi.org/10.5281/zenodo.19286845
Focus: How governance evaluation can be externalized and executed consistently across heterogeneous systems before consequential execution.
- Governance control plane and execution-plane separation
- Governance submissions, evaluation pipelines, and execution gates
- Lifecycle state and governance ledger
- Binding governance decisions to operational outcomes
Read the Paper
Study the architecture that provides the broader framework for deterministic runtime governance.
Open ReadingCompanion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Confirm your understanding of the governance and execution-plane architecture.
Take the QuizAGCP: A Deterministic Execution-Layer Governance Control Plane
For Autonomous and Programmatic Systems
Willis, J. M. (2026). AGCP: A Deterministic Execution-Layer Governance Control Plane for Autonomous and Programmatic Systems. Sustainable Future Tech, Inc. https://doi.org/10.5281/zenodo.19323672
Focus: The concrete AGCP control-plane model for governing state-changing actions before they become operationally reachable.
- Governed action proposal and action envelope
- Five-stage pipeline: validate, evaluate, authorize, and commit
- Structural refusal and append-only governance ledger
- Ledger-derived lifecycle state and commit-bound execution
Companion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Test your understanding of AGCP’s deterministic pipeline, artifacts, lifecycle, and commit semantics.
Take the QuizFormal Execution Semantics and Safety Invariants
For Governance Control Planes in Autonomous Systems
Willis, J. M. (2026). Formal Execution Semantics and Safety Invariants for Governance Control Planes in Autonomous Systems. Sustainable Future Tech, Inc. https://doi.org/10.5281/zenodo.19241732
Focus: The formal lifecycle, ordering, determinism, and invariant properties needed to make runtime governance inspectable, testable, and replayable.
- State-transition systems and lifecycle semantics
- Safety invariants and impossible-transition protections
- Deterministic ordering and decision consistency
- Replayability as a governance assurance property
Read the Paper
Study the formal execution model and the safety properties underlying deterministic governance control planes.
Open ReadingCompanion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Confirm your understanding of lifecycle states, invariants, ordering, and replay.
Take the QuizAGCP Operating Model and Semantic Architecture
Proposal Schemas, Governance Context, Canonical State, and Commit-Bound Execution Semantics for Autonomous Systems
Willis, J. (2026). AGCP Operating Model and Semantic Architecture: Proposal Schemas, Governance Context, Canonical State, and Commit-Bound Execution Semantics for Autonomous Systems. Sustainable Future Tech, Inc. https://doi.org/10.5281/zenodo.20330062
Focus: The semantic contract connecting a proposed action, governance context, qualified canonical state, governance decision, and commit.
- Proposal schema and governance context envelope
- Canonical-state qualification
- Evidence qualification and commit-bound semantics
- Governance receipts and portable decision evidence
Read the Paper
Study the semantic distinctions that keep proposals, context, authority, state, decisions, and execution from collapsing into a single workflow.
Open ReadingCompanion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Test your understanding of proposal semantics, canonical state, governance context, and receipts.
Take the QuizThe PBSAI Governance Ecosystem
A Multi-Agent AI Reference Architecture for Securing Enterprise AI Estates
Willis, J. M. (2026). The PBSAI Governance Ecosystem: A Multi-Agent AI Reference Architecture for Securing Enterprise AI Estates. arXiv:2602.11301. https://arxiv.org/pdf/2602.11301
Focus: How a multi-agent cybersecurity ecosystem can organize roles, shared context, structured outputs, and evidence while keeping execution-layer governance distinct.
- Enterprise AI estate and twelve-domain taxonomy
- Bounded agent families and shared context envelopes
- Structured output contracts and governance evidence
- Separation of agent orchestration from runtime execution governance
Read the Paper
Read the multi-agent reference architecture that provides the enterprise cybersecurity context for the runtime-governance corpus.
Open ReadingCompanion PowerPoint
Open or download the PowerPoint presentation that accompanies this paper and video briefing.
Open PowerPointCheck Your Knowledge
Confirm your understanding of the PBSAI ecosystem, bounded agents, structured outputs, and governance architecture.
Take the QuizAGCP Runtime Governance Assessment Readiness Check
After completing the reading series, use the readiness check to characterize the governance controls you currently have and how far those controls reach toward consequential execution.
The readiness check is a pre-assessment diagnostic, not an AGCP conformance assessment. It is intended to distinguish conventional model safeguards, system guardrails, permissions, approvals, enforcement surfaces, risk re-evaluation, governance compilation, and action-specific runtime governance.
Return to the AI Runtime Governance Learning Guide
After completing all six papers, companion briefings, and knowledge checks, return to the AI Runtime Governance Learning Guide for an integrated review of the series. The guide brings the six readings together and shows how their lessons connect across the execution-governance problem, runtime governance architecture, the AGCP control plane, formal safety semantics, operating semantics, and multi-agent enterprise application.