CouncilReview
Private pilot

Pressure-test your
proposal
before it's judged.

CouncilReview runs your draft through a panel of adversarial judges and a fixed, psychometrically-designed rubric, then tells you whether it's ready to send — and exactly where it isn't.

Join the waitlistSee how it works

No score. No rewrite. A verdict you can act on.

Proceed with Caution

Clear plan, soft on the ask

The proposal explains the solution well, but it doesn't yet make the case for why this approach, at this price, is the one to approve now.

Reasoning
Evidence
Audience
Adversarial
Completeness
Presentation
Held back by Structural Validity · Evidence Quality

Your reader is looking for a reason to say no. CouncilReview is built to be that reader — before you face them.

High-stakes writing gets one shot in front of someone who decides. Most AI is trained to be on your side — to help, to agree, to make it sound good. CouncilReview is built to do the opposite: to be the hostile reader you're about to face, before you face them.

How it works

Four readers, each trying to break it.

First it diagnoses the draft — audience, goal, and stakes — so everything after is judged against the right bar. Then a council of adversarial readers goes to work.

R

Reasoning

Tests whether the logic holds — whether each claim actually follows from what comes before it.

E

Evidence

Asks what's actually backed up, and where assertions stand in for proof the reader will want.

A

Audience

Reads as the person who decides — whether the draft lands for them, at their level, with their priorities.

S

Skeptic

A dedicated adversary — hunts the obvious objection you didn't address and the weakest seam to attack.

→ then

Disagreement is kept, not smoothed

Findings reconcile into one report — where the judges disagree, the tension is shown rather than averaged away.

→ and

Pinned to your own words

Every weak point is named and anchored to the exact line in your draft that triggered it.

→ you get

A readiness verdict

A clear call — ready, or not yet — with what to shore up before you send.

The rubric is fixed and psychometrically designed. It doesn't change from document to document — which is what lets it read the same way every time.
The report

What comes back

A readiness verdict, the weak points pinned to your text, and what to shore up before you send. A sample report:

Proceed with Caution

Clear plan, soft on the ask

The proposal explains the solution well, but it doesn't yet make the case for why this approach, at this price, is the one to approve now.

The scope and timeline are easy to follow and the writing is clean. What holds it back is that the central recommendation is asserted more than justified: the reader can see what you propose, but not why it beats the alternatives they'll have in mind, or why the cost is the right one to commit to today.

Held back by: Structural Validity · Evidence Quality · Unaddressed Obvious Challenge
We read this as an advocacy draft addressed to a client decision-maker, aiming to secure approval for a software rollout and its budget. Near-final draft.
Reasoning
Evidence
Audience
Adversarial
Completeness
Presentation

Where it cracks

The recommendation outruns its justification
From your draft
“We recommend a phased rollout beginning with the East regional warehouse.”

The draft states the phased approach as the plan, but never says why phasing beats a single cutover, or why East is the right place to start. A reader weighing risk will ask both.

Recommendation

Add one or two sentences that give the reason: what phasing buys (less disruption, a contained test) and why East specifically (lowest volume, most flexible staff). The gap is a missing “why,” not missing detail.

The price arrives without a frame
From your draft
“Total implementation is priced at $148,000, with annual support at $22,000.”

The numbers are concrete but stand alone. The reader gets the figure before any sense of what it buys relative to the current manual process or a competing bid.

Recommendation

Anchor the figure to something — the cost of the status quo, the payback period, or what's included. One sentence turns a number into an investment case.

Before you send

Confirm: Confirm the East warehouse volume figures.

The phasing rationale rests on East being the lowest-risk starting point; if the volume numbers are off, the recommendation shifts.

The line we hold

What CouncilReview is — and isn't

It is

  • +A pressure-test you run before you send
  • +A repeatable readiness verdict, with gaps pinned to your text
  • +Private by design

It isn't

  • −A grader handing out a score
  • −A ghostwriter that rewrites for you
  • −A tool that tries to please you
Private pilot

In active private pilot

The method is built and runs the same way every time. What we're doing now is validating and tuning it against real documents and reviewer outcomes. Join the waitlist and we'll reach out as we open spots.

Private by design. Your draft is sent to AI providers to generate the review, then discarded. We don't store your draft or your full report — only run metadata, to keep improving the instrument.

Join the waitlist

We use this only to contact you about the pilot.

Pressure-test important thinking before you act on it.