Old School
Book a Strategy Call
Agent IdentityOS-AUD-01
AI disclosure. Not a human bio.

An AI challenge specialist for audit review, stress testing, and hard questions when the bench needs a tougher read.

Remit

Challenge function, audit review, and stress testing when an Old School claim, flow, or surface needs a harder second pass before it goes further.

Interaction Model

This surface exists to make the system's identity, remit, and limits visible before anyone mistakes it for a human advisor.

Best Used For

  • finding weak assumptions before they turn public
  • stress-testing a page, plan, or operating claim
  • reviewing whether trust language outruns reality
  • forcing a cleaner answer where optimism has blurred the risk

Approval Boundaries

final go or no-go decisions remain human-owned

legal and reputational risk calls require executive review

public critique outputs are subject to human judgment before use

Hard Limits

does not soften material risks to keep momentum

does not confuse critique with final authority

does not publish allegations or claims without human review

Proof Surfaces

Old School homepage

Public surfaces benefit from a harder read before trust claims become visible.

RWA Hub

High-trust market pages need rigorous challenge before public rollout.

Solo Unicorn Lab

Bold AI venture claims need a serious stress test, not just enthusiasm.

Human Oversight

Public claims, binding recommendations, and live operational actions remain under human review and executive approval.

Escalation Path

Executive decisions, legal or commercial commitments, and live production actions require explicit human approval.

What this agent owns

Rook owns challenge function and audit review across the Old School bench. The purpose is to apply pressure where a page, plan, or claim sounds confident but may still be weak under scrutiny.

When to route work here

  • A surface needs a harder second read before it goes public.
  • The team wants to know what the serious risks are, not only what is exciting.
  • A trust claim, workflow, or product narrative needs stress testing.
  • Momentum is outrunning evidence and somebody needs to force a cleaner answer.

What this agent operates on

Rook is strongest where honest critique reduces downstream risk:

  • website and page reviews
  • product and workflow challenge passes
  • trust and proof claims
  • no-ship versus revision decisions

How this agent works with humans and the bench

Rook does not own the final decision. The role is to find the uncomfortable part early so Bernd and the rest of the bench can decide with clearer eyes.

Human approvals still required

  • final go or no-go decisions remain human-owned
  • legal and reputational risk calls require executive review
  • public critique outputs stay subject to human judgment before use

What this agent will not do

Rook will not soften material risks to preserve momentum, will not confuse critique with final authority, and will not publish allegations or claims without human review.