CONTEXT(CLONE)
Sign in

Review a large services agreement by legal, finance, security, privacy

For: Lead counsel on an outsourcing, cloud or managed-services agreement with many schedules

Pattern: Specialist panelNeeds live modelsDesigned for 8 to 200 agents
Get early accessSign in to use

The pain today

The agreement and schedules run to hundreds of pages. Legal reads the front end while service levels, pricing mechanics, security and data processing schedules get a late look from busy specialists, and conflicts between schedules slip through.

The ask

“I attached the draft services agreement with all its schedules and our internal requirements for legal, finance, security and data protection. Review each part from each of those angles, tell me what fails our requirements, and point out where one schedule contradicts another.”

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

  • Draft master agreement and all schedules
  • Internal legal, finance, security and privacy requirements
  • Service description and pricing model
  • Previous version or markup, if any

The unit of work

One worker task per one schedule section under one reviewer brief.

Why a swarm fits

Each section is read against one brief's checklist, which keeps context small. The risk hides between schedules, so many scoped readings are followed by a cross-schedule reconciliation.

Not for

A short standard agreement with no schedules: one strong model with your checklist covers it in a single pass.

The decision tree

5 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 schedule section contain obligations, prices, service levels, security or data terms, as opposed to definitions or repeated boilerplate?

    Sees only: One section's heading and first paragraphs

    Why: Keeps specialists' briefs on the sections that can hurt.

    • Yes: 0.60 or higherthenAccept
    • Unsure: 0.30 up to 0.60thenAccept
    • No: below 0.30thenSkip this unit

  2. Before workers, before a task runs

    Small worker or strong modelYes or no, with a probability

    Before a task runs

    Does this section contain a formula, a service credit table or a technical security standard that the brief's checklist cannot test by reading alone?

    Sees only: The section text and the brief's checklist items for it

    Why: Pricing mechanics and credit tables are where small readers go wrong.

    • Yes: 0.60 or higherthenEscalate to a strong model
    • Unsure: 0.35 up to 0.60thenEscalate to a strong model
    • No: below 0.35thenAccept

  3. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Does the quoted draft text fall short of the internal requirement cited, for example a recovery time longer than our continuity standard allows?

    Sees only: The quoted draft text and the quoted internal requirement

    Why: A finding must show both sides so the specialist can confirm it at a glance.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenEscalate to a strong model
    • No: below 0.50thenReject and retry

  4. Reconciler, while merging

    Conflict checkA choice among options

    While reconciling

    Do these two schedules set different values for the same thing, such as service hours, liability for data loss or price indexation?

    Sees only: One passage from each schedule and the order-of-precedence clause

    Why: The risk in long agreements hides between the schedules.

    • Different values, precedence settles itthenAccept
    • Different values, precedence unclearthenMark unresolved
    • Different subjectsthenAccept

  5. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Would accepting the draft as it stands leave a requirement owned by legal, finance, security or privacy unmet?

    Sees only: The issue, the requirement and its owner

    Why: Each specialist owner decides whether to push, trade or accept.

    Accountable: Lead counsel and each specialist owner decide which issues to push, trade or accept before signature.

    • Yes: 0.40 or higherthenAsk a person
    • Unsure: 0.15 up to 0.40thenAsk a person
    • No: below 0.15thenAccept

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 maps every section to the briefs that must read it and lists cross-schedule dependencies.

    Decisions here:1. Scope check

  2. Workers

    Mid-sized workers, each holding one brief (legal, finance, security, privacy, operations), review one section against its checklist.

    Designed for 8 to 200 agents, one worker task per one schedule section under one reviewer brief. Each worker receives only its own unit.

    Decisions here:2. Small worker or strong model

  3. Judge, from a different model family

    A decision model from a different family checks that each finding is supported by the quoted text and the cited requirement.

    Decisions here:3. Evidence check

  4. Reconciler

    A strong reasoning model joins findings, surfaces contradictions between schedules and orders issues by brief and severity.

    Decisions here:4. Conflict check

  5. Accountable person

    Lead counsel and each specialist owner decide which issues to push, trade or accept before signature.

    Decisions here:5. Person decides

Checked before anything is accepted

  • Each finding quotes the draft text and the internal requirement it fails
  • Service credits, caps and pricing formulas are recomputed from the schedule's own examples
  • Order of precedence is applied before two schedules are reported as conflicting
  • Requirements with no matching provision are listed as gaps

What comes back

  • Issues list per brief with clause references
  • Contradictions between schedules and the main terms
  • Internal requirements the draft does not address
  • Questions for the supplier and for internal specialists

What to measure

  • Issues specialists confirm as real
  • Cross-schedule conflicts found before signature
  • Specialist hours spent per review round

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

Draft and rank fallback wording for contested clauses

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

Counsel drafts one or two alternatives under time pressure and sends the first that reads well.

Pattern: TournamentNeeds live models6 decisionsDesigned for 8 to 150 agents