Productive Value–Productive Power (PV-PP) Framework

A formal framework for analyzing decision, productive capability, viability, governance, selection, and action under constraints.
Research program developed by Lance Amundsen

About the PV-PP framework

The Productive Value–Productive Power (PV-PP) framework is a research program for representing decisions and interactions in terms of what agents can produce, preserve, exchange, lose, restore, and realize.

At its foundation, Productive Value (PV) concerns value conveyed through goods, services, information, currency, and other exchanges. Productive Power (PP) concerns the capacity to generate or provide Productive Value. The framework extends these concepts into a structured decision architecture involving actual and perceived state, viability, governing domains, policy construction, constraints, restoration adequacy, selection, execution, and state transition.

The framework is intended to preserve structure that can disappear when heterogeneous consequences are compressed prematurely into a single scalar objective. It therefore represents feasibility, viability, dependencies, recovery paths, and decision boundaries explicitly where the problem requires them.

Framework Architecture

10 — Core Framework

Layer 1 and Layer 2 architecture, state and transition semantics, perception interfaces, graph structure, and supporting specifications.

Browse Core Framework

15 — Operators

Current operator-owner specifications and selected supporting specifications governing the PV-PP decision architecture.

Browse Operators

20 — Stack and Layer Governance

Ownership maps, stack references, invariants, architecture diagrams, reader guidance, and interface governance.

Browse Governance

30 — Applications and Extensions

Selected Layer 3 applications and Layer 4 scaling and research extensions, with their canonical, provisional, or exploratory status preserved.

Browse Applications

40 — Runtime and Execution Formalization

The public architecture connecting framework governance to bounded runtime interfaces and tool-action admission.

Browse Runtime Architecture

Formal Proof and Published Benchmarks

Scalar Reduction Proof Program

Formal research examining the conditions under which PV-PP decision structures can, and cannot, be represented through scalar aggregation.

Proof Repository

Project Page

PV-PP Agent Decision Layer Demo

Enterprise-compute incident benchmark comparing scalar optimization against corridor-preserving PV-PP decision logic.

Repository

Project Page

Grenade Self-Sacrifice Benchmark

Extreme-threat benchmark examining self-sacrifice and preservation of other agents under severe constraints.

Repository

Project Page

AI Gridworld Safe Benchmark

Goal-seeking benchmark involving hazards, traps, unsafe shortcuts, and governance constraints.

Repository

Project Page

Authority and Research Status

This repository is a curated public release of a substantially larger research program. Public availability does not by itself make a document canonical.

Current owner specifications and governance documents control within their declared scopes. Supporting, provisional-canonical, exploratory, application, example, benchmark, and simulation materials retain their stated status.

Public does not mean canonical.
Publication status and architectural authority are separate.

Read the Authority and Status rules

Scope of This Public Release

The internal PV-PP research tree contains substantially more material than is presented here. Exploratory research, unpublished papers, submitted manuscripts, books, intellectual-property work, developmental simulations, implementation prototypes, internal runtime materials, publication workspaces, and project-management files are intentionally excluded.

Specialized research programs that have been independently published are referenced from this repository rather than duplicated. This preserves a single authoritative public source for each project and reduces version divergence.

Read the August 2026 Release Notes