Skip to content
PROJECT SHADOW 1.0.1 · CORRECTED R1 REFERENCE · PRELIVE · 2026-08-17

Project Shadow 1.0.1 contains no Myth package. Generic Myth v0.2.0 and Full-Canon Myth v0.3.5 are separate optional companions; both default off, neither is required by R1, and neither can authorize action or change an R1 result. PBHP remains in active testing; publication is not certification, independent validation, or deployment authorization.

A—ZTerms with operational consequences

The language of the pause.

These terms are short because they must survive pressure. Each names a state, check, or governance object; none is a slogan that authorizes action by itself.

LONG-FORM COMPANION

The field manual keeps the explanation visible.

Read the argument, numbered procedure, worked cases, failure contrasts, and evidence state behind this chamber.

Read the full chapter →
A—G

Action through Gap

The beginning of the decision contract.

ACTION

Canonical action

Who will do what, to whom, with which tools and data, under what authority, deadline, and consequence.

Material changes reopen classification. A vague goal is not an action.

CAPA

Corrective and Preventive Action

An owned loop for root cause, correction, prevention, verification, and evidence-based closure.

CAPA addresses system defects, not just individual blame.

CLA

Context Load Audit

A check on whether the current evaluator can reliably preserve and judge the decision-bearing context.

Critical load at critical stakes binds to refresh, routing, or refusal.

DOOR / WALL / GAP

Path classification

Door is permitted within conditions; Wall is prohibited; Gap is genuinely unresolved.

A Gap remains a first-class state. It is not a temporary Door created by deadline pressure.

FPV

False Positive Validation

A structured challenge to an unnecessary pause that must supply an alternative safe Door.

Repeated rejected challenges can become a wear-down signal.

H—R

Harm through Receipt

The objects that bind uncertainty to accountable action.

HARM FLOOR

Worst binding state

The most restrictive credible harm, power, irreversibility, or evidence state that controls the action.

It cannot be averaged away by favorable readings elsewhere.

IAM

Inference & Attribution Module

Separates content, accuracy, negligence, recklessness, and knowing deception.

Intent attribution requires a higher evidence trail than factual correction.

LOCK / FLOOD

Reasoning collapse modes

LOCK ends inquiry too early; FLOOD avoids commitment through endless inquiry.

Both recover through anchored, bounded, reversible action and explicit review.

MAYBE / THEREFORE

Dissent before bounded conclusion

Maybe carries the strongest honest objection; Therefore states only what may proceed despite it.

The Therefore does not erase or defeat the Maybe.

OQL

Open Question Ledger

A durable record of consequential unknowns, owners, dependencies, dates, and closure rules.

Unknown remains unknown until evidence or redesign closes it.

RECEIPT

Write-ahead decision record

A reconstructable binding of action, evidence, unknowns, dissent, gate, Door, owner, and review trigger.

Integrity and completeness do not prove truth or correctness.

S—Z

Sovereignty through zero contempt

The boundaries that keep the protocol from becoming its own harm.

SOVEREIGNTY

Human responsibility

The protocol introduces friction and constraints without inheriting moral or legal ownership of the decision.

Accountable actors remain accountable.

TIER

Required rigor

HUMAN, MIN, CORE, or ULTRA depth matched to consequence and uncertainty.

A higher tier adds evidence and review; it never relaxes a floor.

WHO PAYS FIRST

Least-powerful-first ordering

Identify the person who absorbs the earliest harm and has the fewest options if the decision is wrong.

Their ability to consent, exit, appeal, recover, and be represented affects gate and Door quality.

ZERO CONTEMPT

Clarity without dehumanization

Name harm and power plainly without inventing motives, humiliating people, or turning them into disposable enemies.

Tone is part of epistemic quality because contempt corrupts stakeholder modeling.