How it works

Execution is not truth.

Truth Engine separates what software was allowed to do, what it attempted, what evidence exists, and what it is justified in claiming now.

The chain

From authority to justified truth.

Each step answers a different question. None can safely be collapsed into the next.

AUTHORITY
→
PERMITTED EXECUTION
→
OCCURRENCE HISTORY
→
EVIDENCE
→
PROOF
→
ADMISSIBILITY
→
CURRENT JUSTIFIED TRUTH

Four distinctions

Do not collapse the steps.

Claim ≠ Truth

A system statement is not automatically justified.

Attempt ≠ Occurrence

Sending an action does not prove the external effect occurred.

Acknowledgement ≠ Proof

A provider response can be evidence without being sufficient proof.

HTTP failure ≠ non-occurrence

A failed or missing response cannot safely prove that an external effect did not happen.

Truth states

Exactly three public states.

SATISFIED

Available evidence justifies that the intended effect occurred.

NOT_SATISFIED

Available evidence justifies that the intended effect did not occur.

UNKNOWN

Available evidence does not justify either certainty.

Reconciliation may still return UNKNOWN.

Flagship case

The provider acted. The response disappeared.

That is where naive retry logic becomes dangerous.

01Action sent

The external provider performs one harmless synthetic effect.

02Response lost

The application cannot safely prove whether the effect occurred.

03UNKNOWN

The engine preserves uncertainty.

04Retry denied

Retry intent is not retry authority. No second effect is sent.

05Reconcile

Read-only observation finds provider evidence.

06SATISFIED

Truth changes only after evidence justifies the claim.

Test it

Run the difficult case yourself.

The Playground is engine-backed. Try retrying uncertainty, then reconcile the evidence.