Proofs

Zeq's doctrine is one line: never say "trust me." Every result the framework produces carries something you can check for yourself — the equation it used, the constants it bound, its uncertainty, and a cryptographic ZeqProof any node can re-verify. Trust is never asked for; it's always demonstrable.

Every compute returns a signed envelope carrying:

$$\text{ZeqProof} = \text{HMAC}(\text{operator} \mid \text{result} \mid \text{zeqond})$$

Because the proof binds the operator, the result, and the exact tick together, a verifier re-runs the maths and re-checks the stamp independently — it never has to trust the node that produced the number.

The layers of trust

Per-result

The ZeqProof binds one computation. Any node can re-verify it offline from the envelope alone.

Per-history

A machine's whole audit chain is sealed so rewriting the past costs real sequential time — tamper-resistant, not just tamper-evident.

Per-precision

Every result is checked against the ≤0.1% bound and independently re-computed before it's returned.

Explore

The envelope

What ships with every result, and how a stranger checks it.

• The Zeq envelope — every field and what it binds
• The verification envelope — re-verify without trusting the source

Precision & the seal

Why the number is accurate and the history holds.

• Precision & proof — the ≤0.1% bar, verdicts, re-evaluation
• Proof of Elapsed Zeqonds — the seal spine that resists rewriting

Prove the whole node

Verify a node, and check quantum outputs.

• Conformance & verification — prove a node's physics is unmodified bit-for-bit, plus honestly-bounded quantum-output verification

The audit chain these proofs live on is part of your machine; the verification APIs are in the reference.