RTT Evaluation Framework

The RTT Evaluation Framework provides a regime‑aware method for determining whether a governance decision, system, or intervention should proceed, adapt, pause, or be contained. It integrates awareness, stress testing, failure mapping, and sufficiency checks into a single evaluative posture.

RTT does not optimize outcomes. It tests survivability across regimes.

What RTT Is#

RTT stands for Regime‑Translation Testing.

It evaluates whether an action:

• Remains coherent when context shifts.
• Preserves invariants under stress.
• Avoids known failure modes.
• Maintains reversibility and legibility.
• Can be interrupted without domination.

RTT is a filter, not a validator of intent.

Why RTT Is Necessary#

Most governance failures occur because decisions are evaluated only where they are made.

RTT exists to prevent:

• Local success becoming global harm.
• Tool overextension across regimes.
• Moral certainty masking structural fragility.
• Speed replacing understanding.

RTT slows decisions just enough to prevent irreversible error.

Core RTT Questions#

Every evaluation asks five questions, in order:

1. What regime is this decision operating in?
   Identify domain, scale, authority structure, and time horizon.

2. What assumptions does it rely on?
   Make implicit constraints explicit.

3. What happens when those assumptions fail?
   Map failure modes and escalation paths.

4. Can the decision be interrupted or reversed?
   Test for phase lock and irreversibility.

5. Does it remain minimally sufficient across regimes?
   Remove excess structure and re‑test.

If any question fails, the decision must adapt or pause.

RTT Evaluation Outcomes#

RTT produces one of four outcomes:

• Proceed — Decision remains coherent across tested regimes.
• Adapt — Decision requires contextual modification.
• Pause — Uncertainty or risk exceeds acceptable bounds.
• Contain — Decision must be isolated to prevent harm.

Proceeding without passing RTT is a governance failure.

RTT and Authority#

RTT does not grant authority. It constrains authority.

Authority may:

• Act only after RTT evaluation.
• Be limited by RTT findings.
• Be paused by RTT uncertainty.
• Be overridden by invariant violation.

RTT protects systems from confident error.

Role of AI in RTT#

AI may assist RTT by:

• Simulating regime shifts.
• Surfacing assumption dependencies.
• Identifying nonlinear effects.
• Highlighting confidence collapse.

AI must not:

• Decide outcomes.
• Authorize action.
• Resolve tradeoffs.
• Mask uncertainty.

RTT remains a human responsibility.

Failure Mode#

RTT fails when:

• Evaluation is rushed.
• Assumptions remain implicit.
• Authority bypasses testing.
• Speed is rewarded over coherence.

At that point, governance learns through damage.

RTT is not bureaucracy.
It is the minimum pause required to avoid irreversible harm.

Decisions that cannot survive RTT should not be scaled.