Skip to content
ChangeSteward

ChangeSteward Risk Scoring™ /

Know who's exposed. Before you sign.

Every clause in your contract, read and scored for both sides. Not a summary — a number, with the reason behind it, on the exact clause causing it.

You (Humpty Dumpty)

0

Elevated

0 Protected100 Exposed

Them (King Mortimer)

0

Protected

0 Protected100 Exposed

Same contract. Two different numbers. That gap is the negotiation, and most teams walk into it without knowing it exists.

How it reads a contract /

Clause by clause, not page by page.

A contract isn't risky. Particular clauses are. The engine goes through each one and works out what it actually obliges you to do, what discretion it hands the other side, and what you can do if it goes wrong.

StewardGraph™ · reading clauses /0 / 6
  1. 01Definitions & Interpretation12
  2. 02Scope of Services34
  3. 04Fees & Payment Schedule48
  4. 05Service Levels (SLA)29
  5. 07Limitation of Liability72
  6. 08Termination41

Clause 07is where you are most exposed. Cap it at 12 months’ fees and your score drops to 28.

Four things, on every clause

  • What you must do. Obligations you’ve taken on, and how hard they are to meet.
  • What they can decide. Discretion handed to the other side — the words “at its sole discretion” are expensive.
  • What you can do about it. Whether you have a remedy when things go wrong, or just a strongly worded email.
  • What it could cost. Where your exposure is capped, and where it isn’t.

Why it’s ours /

StewardGraph™ doesn’t read a clause on its own. It reads it against every variation that came after it, against the other clauses it interacts with, and the engine's estimate of where this sits against comparable deals. A liability cap means something different when the indemnity beneath it is uncapped — and that’s the sort of thing a summary tool misses.

The proof /

We show our working.

0–100

per party

Scored separately for you and for them. The same clause is rarely fair to both.

Every

score cites its clause

Each factor names the clause it came from and what it added or removed.

Shown

not hidden

The engine's confidence appears on every score, and the engine version is recorded against it.

0.98

verbatim threshold

Before a deed can be sent, every accepted passage is machine-checked against it. One failure blocks the send.

It re-scores itself the moment anything moves.

Suggest a price change and the score updates before you’ve sent it. You find out what a change costs you while you can still take it back — not six months later when someone invokes the clause.

And when the engine spots a fix, it tells you what the fix is worth. Not “consider capping indemnity” — cap it at 12 months and this number goes from 41 to 28.

Wall Works Agreement — risk
Risk analysis showing per-party scores, category breakdowns and Steward's recommendation

One more thing /

You decide who sees it.

Your score is yours. Theirs is theirs. Neither side sees the other’s by default, because handing the other party a map of where you’re weak is not a feature.

If you’d rather they couldn’t run the analysis at all, you can switch it off for them — a one-off $5 for that contract. We will tell them it’s been switched off, and who did it. Quietly blinding your counter-party isn’t something we’re willing to help with.

Get started /

Find out what your contract actually says.

Send us one messy contract — the one with all the amendments nobody can keep straight. We'll show you what it says today and where you're exposed.