The build/review loop

AI Factory

AI Factory runs every change through a Prime Builder and a Loyal Opposition – two independent AI roles that build and adversarially review each other's work – before a deterministic gate and a human make the final call.

What it isAn adversarial build-and-review loop
RolesPrime Builder · Loyal Opposition · human gate
Arbitrated byA deterministic, non-LLM gate
StageWorking prototype
01The outcome
Before

Building a governed agent system means one model produces a change and a human reviews every one for correctness and safety – or, worse, nothing ships on more than a single model's say-so.

After

Every change is built by a Prime Builder and adversarially reviewed by an independent Loyal Opposition before a human signs off – so nothing ships on one model's judgement alone, and the human reviews a decision, not a blank page.

AI Factory is how the stack builds more of itself safely: governed systems generated and checked under the same identities and audit trail everything else uses.

02How it works

The idea is simple and the discipline is the point: no change is trusted on one model's word. A builder proposes, an opponent tries to break it, a deterministic gate decides, and a human has the last word.

ROLE 01

Prime Builder

Produces the change – the artefact under review – bound by a versioned role constitution that says what "good" means for this task.

ROLE 02

Loyal Opposition

An independent reviewer whose job is to find what's wrong: correctness, safety, and adherence to the same constitution, argued explicitly.

THE GATE

Deterministic arbitration

A pure, non-LLM gate arbitrates the two into one outcome – GO, REVISE, NO‑GO, or HUMAN‑ESCALATED. Same inputs, same decision, every time.

THE HUMAN

The final gate

A human makes the call the machine escalated – reviewing an argued recommendation and its dissent, not an unexamined diff.

Why this shape The whole point of the stack is verifiable trust, so the tool that builds on it holds itself to the same bar. Every handoff is a VAID-signed artefact on a tamper-evident chain: you can see which role produced what, and why the gate decided as it did. This very site was built through that loop.
03Proof

The gate is deterministic and pure – no model in the arbitration path – so the same builder output and the same review always produce the same decision. That's what makes the record trustworthy.

Live demo · the real build → review → gate loop
Autoplay

Watch a real build–review–gate cycle run

A recorded Prime Builder / Loyal Opposition campaign, arbitrated by the deterministic gate — start to finish.

~1.9 MB · loads on tap

Prefer a guided version? Walk through it in presenter mode →

Example · a gate decision
build → review → gate
prime_builder   → change: add rate-limit to /mint
loyal_opposition→ MAJOR: no test for the burst case
                  recommended_decision: REVISE
────────────────────────────────
gate (deterministic) → REVISE
  handoff: VAID-signed, on the audit chain
  human: reviews the dissent, not a blank diff

Two independent AI roles, one deterministic gate, a human at the end – every step signed and on the record.

SYNTHERA is the trust layer for multi-agent systems: every agent gets a verifiable identity, scoped authority and a tamper-evident record, so software from different teams, vendors and frameworks can act on each other’s behalf without custom glue between every pair.

Talk to us about AI Factory.

Talk to us about AI Factory