CONTEXT(CLONE)
Sign in

Generate and rank deal protections for each diligence finding

For: Deal counsel or corporate development lead turning findings into contract terms

Pattern: TournamentNeeds live modelsDesigned for 8 to 250 agents
Get early accessSign in to use

The pain today

Each red flag needs a response: a specific indemnity, a price adjustment, an escrow, a condition before closing or a covenant. Under time pressure every finding gets the same indemnity wording, whether or not it fits the risk.

The ask

“I attached the due diligence report, the draft share purchase agreement and a note on what the seller has already refused. For each finding, propose several ways to protect us, draft the wording, rank the options by how well they cover the risk and how negotiable they are, and explain the ranking.”

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

  • Due diligence report with findings and sources
  • Draft share purchase agreement
  • Negotiation history and seller positions
  • Deal team's risk appetite and priorities

The unit of work

One worker task per one candidate protection for one finding.

Why a swarm fits

A report holds many findings and each admits several kinds of protection. Candidates are cheap and independent, and each is scored only against one finding and the clauses it touches, so they narrow in rounds.

Not for

A deal with a few findings and warranty insurance covering them: counsel's standard approach is quicker.

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 checkA choice among options

    Before work starts on a unit

    Is the risk in this finding already covered by a warranty, indemnity or price mechanism in the current draft?

    Sees only: The finding's risk statement and the draft clauses retrieved for it

    Why: No options are drafted for risks the agreement already handles.

    • Fully coveredthenSkip this unit
    • Partly coveredthenAccept
    • Not coveredthenAccept

  2. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Would this wording respond if the finding's risk scenario occurred, given the caps, baskets and time limits in the draft that apply to it?

    Sees only: The candidate, the risk scenario and the draft's limitation clauses

    Why: An indemnity swallowed by the general cap protects nobody.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenEscalate to a strong model
    • No: below 0.50thenSkip this unit

  3. Planner, while planning

    Scope checkA score

    Before work starts on a unit

    How much of the loss described in the finding's scenario would this protection recover for the buyer?

    Sees only: The candidate and the finding's loss scenario

    Why: Ranks by fit to the risk, not by how familiar the wording is.

    • 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

    Given the positions the seller has already refused, how negotiable is this protection?

    Sees only: The candidate and the negotiation history note

    Why: Keeps the shortlist to asks that can still be won.

    • Not yet refused in any formthenAccept
    • A softer form of a refused askthenContinue
    • Already refused outrightthenSkip this unit

  5. Run control, between rounds

    Another round?Yes or no, with a probability

    Between rounds

    Do the surviving candidates for this finding use different protection types, so that another round could still reorder them?

    Sees only: The surviving candidates for one finding

    Why: Stops when only variations of one idea remain.

    • 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

    Would this protection change the price, the escrow or the liability regime put to the seller?

    Sees only: The finalist and its effect on the draft's liability clauses

    Why: Deal counsel and the deal lead choose what to ask for and what to trade.

    Accountable: Deal counsel and the deal lead choose which protections to ask for and what to trade; the swarm drafts options only.

    • 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 each finding's risk: what could go wrong, when, and what the draft already covers.

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

  2. Workers

    Small fast workers from several families each draft a different protection type for one finding.

    Designed for 8 to 250 agents, one worker task per one candidate protection for one finding. Each worker receives only its own unit.

  3. Judge, from a different model family

    Judges from families that did not draft score risk coverage, enforceability gaps and negotiability, round by round.

    Decisions here:2. Evidence check

  4. Reconciler

    A strong reasoning model presents finalists per finding and the combined effect on caps, baskets and time limits.

    Decisions here:5. Another round?

  5. Accountable person

    Deal counsel and the deal lead choose which protections to ask for and what to trade; the swarm drafts options only.

    Decisions here:6. Person decides

Checked before anything is accepted

  • Each candidate is tested against the finding's risk scenario, not against generic wording
  • Interaction with caps, baskets, time limits and disclosure is checked in the draft agreement
  • Defined terms and cross-references must exist in the agreement
  • No model scores a candidate it drafted

What comes back

  • Ranked protections per finding with draft wording
  • Trade-offs of each finalist
  • Combined effect on the liability regime of the agreement
  • Findings where no contractual protection fits and price is the only lever

What to measure

  • Share of finalists counsel takes forward
  • Findings left without a tailored protection at signing
  • Drafting hours per finding

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 Due diligence and M&A

Red-flag customer contracts for change of control and exit rights

For: M&A associate or corporate development manager on a buy-side review

The value of the target sits in its customer contracts, and the data room holds hundreds with order forms and amendments.

Pattern: Map, verify, reduceRuns today6 decisionsDesigned for 4 to 300 agents

Test every claim in the information memorandum against the data room

For: Private equity investment manager or corporate development lead before a binding offer

The seller's memorandum and management presentation make claims about customers, retention, pipeline, capacity and compliance.

Pattern: Adversarial reviewNeeds live models6 decisionsDesigned for 8 to 400 agents