CONTEXT(CLONE)
Sign in

Draft and rank fallback wording for contested clauses

For: Commercial counsel refreshing a template or stuck on a contested set of clauses

Pattern: TournamentNeeds live modelsDesigned for 8 to 150 agents
Get early accessSign in to use

The pain today

Counsel drafts one or two alternatives under time pressure and sends the first that reads well. Options that protect the same interest and are easier for the other side to accept are never written down or compared.

The ask

“I attached our template, the other side's markup, our playbook and a note on what we must protect in each contested clause. For each clause, draft several fallback wordings, rank them by how well they protect us and how likely the other side is to accept, and explain the trade-off.”

Plain words, as you would say it to a colleague. Edit it to fit your case before you send it.

What you attach or connect

  • Template or current draft agreement
  • Counterparty markup and comments
  • Playbook interests and limits per clause
  • Earlier agreed positions with similar counterparties

The unit of work

One worker task per one candidate wording for one clause.

Why a swarm fits

Candidates are cheap to write and independent of each other. Each is scored only against the stated interest, the markup and the defined terms it touches, so many can be produced and narrowed in rounds.

Not for

A clause with one obvious market-standard wording: asking one strong model once is cheaper and just as good.

The decision tree

6 typed decisions, each with an action for every answer

At fixed moments in a run, the engine puts one narrow question to a decision model. The decision model never writes text: it answers yes or no with a probability, picks from listed options, or gives a score, about a small slice of the material. The engine then does exactly what this tree says, which is what makes the run auditable. The thresholds are the template's design values, not measured results.

  1. Planner, while planning

    Scope checkYes or no, with a probability

    Before work starts on a unit

    Does this candidate wording stay inside every hard limit the playbook sets for the clause, such as the minimum cap and the forbidden carve-outs?

    Sees only: One candidate and the playbook's hard limits for that clause

    Why: Candidates that breach a hard limit never reach the scoring rounds.

    • Yes: 0.80 or higherthenAccept
    • Unsure: 0.50 up to 0.80thenEscalate to a strong model
    • No: below 0.50thenSkip this unit

  2. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Does every defined term and clause number used in this candidate exist in the draft agreement shown?

    Sees only: The candidate, the draft's definitions list and its clause numbering

    Why: Invented defined terms make a clause unusable however well it reads.

    • Yes: 0.90 or higherthenAccept
    • Unsure: 0.60 up to 0.90thenReject and retry
    • No: below 0.60thenReject and retry

  3. Planner, while planning

    Scope checkA score

    Before work starts on a unit

    How fully does this wording protect the interest stated for the clause if the loss scenario in the playbook note occurs?

    Sees only: The candidate, the stated interest and the loss scenario

    Why: Narrows the field on substance before anyone weighs style.

    • High: 0.70 or higherthenAccept
    • Middle: 0.40 up to 0.70thenContinue
    • Low: below 0.40thenSkip this unit

  4. Scope checkA choice among options

    Before work starts on a unit

    Set against the counterparty's markup comment, how would this wording land?

    Sees only: The candidate and the counterparty's comment on that clause

    Why: A protective clause the other side already refused is not a fallback.

    • Answers their stated objectionthenAccept
    • Partly answers itthenContinue
    • Repeats what they struck outthenSkip this unit

  5. Run control, between rounds

    Another round?Yes or no, with a probability

    Between rounds

    Do the remaining finalists differ in substance, as opposed to wording only, so that another round could change the ranking?

    Sees only: The finalists for one clause, side by side

    Why: Stops paying for rounds that only polish.

    • Yes: 0.60 or higherthenContinue
    • Unsure: 0.30 up to 0.60thenStop
    • No: below 0.30thenStop

  6. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Does this finalist concede anything compared with our standard position on liability, indemnity or termination?

    Sees only: The finalist and the standard clause

    Why: Counsel owns every concession that leaves the building.

    Accountable: Counsel chooses the wording that goes to the counterparty and owns the legal risk it carries.

    • Yes: 0.30 or higherthenAsk a person
    • Unsure: 0.10 up to 0.30thenAsk a person
    • No: below 0.10thenAccept

The fleet: who does what

Model tiers by role, not brands: you choose the models. Strong reasoning models plan and reconcile, small fast models do the wide work, and the judge is a decision model from a different family, so it does not share the workers' blind spots.

  1. Planner

    A strong reasoning model states, per clause, the interest to protect, the hard limits and what the counterparty objected to.

    Decisions here:1. Scope check3. Scope check4. Scope check

  2. Workers

    Small fast workers from several families each draft candidate wordings using different drafting approaches.

    Designed for 8 to 150 agents, one worker task per one candidate wording for one clause. Each worker receives only its own unit.

  3. Judge, from a different model family

    Judges from families that did not draft score each candidate on protection, acceptability and fit, round by round.

    Decisions here:2. Evidence check

  4. Reconciler

    A strong reasoning model presents the finalists per clause with the trade-offs and any knock-on changes elsewhere.

    Decisions here:5. Another round?

  5. Accountable person

    Counsel chooses the wording that goes to the counterparty and owns the legal risk it carries.

    Decisions here:6. Person decides

Checked before anything is accepted

  • Every defined term and cross-reference in a candidate exists in the agreement
  • Candidates that breach a playbook hard limit are eliminated before scoring
  • Each candidate is checked for conflict with related clauses such as caps and exclusions
  • No model scores a candidate it drafted

What comes back

  • Ranked finalist wordings per clause with reasons
  • What each finalist concedes compared with our standard
  • Knock-on edits needed in other clauses
  • Eliminated approaches and why they failed

What to measure

  • Share of finalists counsel uses with light edits
  • Negotiation rounds per contested clause
  • Drafting hours per clause

Names of measures only. No result is claimed for this template.

Templates open in the workspace chat with the ask filled in. Nothing runs until you send it.

Get early accessSign in to use

More in Legal and contracts

Find where signed contracts deviate from the contract playbook

For: General counsel or legal operations lead who inherited a contract portfolio

Nobody knows which signed agreements carry uncapped liability, odd governing law or weak termination rights.

Pattern: Map, verify, reduceRuns today5 decisionsDesigned for 4 to 250 agents

Settle assignment and change-of-control readings before a transfer

For: In-house or deal counsel planning an intra-group reorganisation, carve-out or sale

Whether consent is needed turns on wording: assignment versus transfer by operation of law, direct versus indirect control, affiliate carve-outs.

Pattern: Cross-examinationNeeds live models5 decisionsDesigned for 6 to 200 agents