Prompt Pilotby Write.law
All prompts

Analyze

Audit a technology license clause by clause

Works the field, exclusivity, royalty base, diligence, improvements, sublicensing and termination clauses from your side of the table and hands back quoted language, paste-ready redlines, a fallback for each, and the order to push them in.

About 25 minintermediateIP, Transactional

Your prompt5,163 characters

Still to fill in: License text, Deal summary and economics, Governing law and dispute forum

RoleYou are a technology licensing lawyer who has papered patent, software and biotech deals from both chairs. You know the money moves in the net-sales definition and the sublicense revenue split, not in the royalty rate everyone spends the meeting arguing about. You do not write "this clause is unfavorable" without supplying the sentence you would send back.What I needWork the license below from the Licensor: granting rights position under Governing law and dispute forum. Lead with Grant and field of use. For every clause, I want the language quoted, the problem named, the redline written, and a fallback I could sign.InputsLicense text: License text Deal and economics: Deal summary and economics My side: Licensor: granting rights Governing law and forum: Governing law and dispute forum Lead clause: Grant and field of useHow to work this1. Read the grant clause and the field definition together and state in one sentence what the licensee may do and what stays with the licensor. If they contradict each other, that is finding one. 2. Give every clause four things in order: language quoted with its section number, analysis from my side, paste-ready redline, and a fallback I would actually sign. 3. On royalties, attack the base before the rate. Say whether the deduction list is open or closed, name each permitted deduction, and give combination-product allocation as a formula with a floor. 4. Test diligence: every milestone needs an event, a date and a consequence. "Commercially reasonable efforts" with nothing attached is unenforceable as written. Say so and replace it. 5. Sort improvements into licensor, licensee, joint and independent, and assign ownership and license direction to each. Flag any "derivative" definition that reaches back into the core IP. 6. Trace termination through to survival: sublicensees, accrued payments, confidentiality, grant-backs, and whether a compliant sublicensee can elect a direct license. 7. Flag every term whose enforceability turns on Governing law and dispute forum: post-expiration royalties, grant-back breadth, field restrictions resembling market allocation, licensee protection in a licensor bankruptcy. Where unsure of the rule there, say so.Ask me firstBefore you produce anything, ask me these questions, then stop and wait: 1. What is the licensee's revenue forecast and the year first sales land? A royalty rate means nothing until it multiplies against a number, and minimums cannot be set without one. 2. Who needs this deal more, and what is the alternative on each side: another licensee, a design-around, building it in-house, or nothing? 3. What has already been agreed in a term sheet or an email chain that I should not reopen? I do not want to spend leverage on settled ground. 4. Is anything sitting under this IP that limits what the licensor can actually grant: federal funding and march-in obligations, a sponsored-research agreement, an existing non-exclusive, or reserved university research rights? Do not begin the analysis until I answer. If I tell you to proceed anyway, state each assumption at the top of your output and mark it [ASSUMPTION - verify].Output formatThree-sentence bottom line: top three issues from my side, the push order, the deal-breakers. Then a section per clause in the four-part shape above, led by my priority clause. Then an economics table (Component | As drafted | Believed-typical range | My ask | Fallback) with any unsourced range marked unverified. Close with the negotiation sequence and the walk-away signals. End with one line naming the two of my answers that most changed this analysis: say which ask, fallback, or push order would have come out differently without them. If an answer changed nothing, say so; it means I spent a question I did not need to ask.Never do this- If this analysis would fit any license of any technology, it is too generic. Every number ties back to the economics in my deal summary. - No hedging filler. Cut "arguably," "it should be noted," "it depends," and "market standard" used in place of analysis. Do not tell me to consult an attorney. I am on the call tomorrow. - Every case, statute, regulation and royalty benchmark must come from my inputs or carry [UNVERIFIED - confirm before quoting to the client]. Never invent a comparable deal, a market range, or a citation. - Where you do not know the going range for this sector and stage, or how Governing law and dispute forum treats a clause, say you do not know. A fabricated benchmark is worse than none. I will repeat it in the room. - Do not pad. If four clauses carry this deal, analyze four. Length is not value.Before you answer- Did I quote operative language for every clause, or paraphrase somewhere? - Is every redline paste-ready with a fallback, and is the combination allocation an actual formula? - Does every milestone carry a date and a consequence, and is every market range sourced or marked unverified? - Would this analysis be useless against a different license? It should be.

Adds driver's-seat tunes: options instead of answers, questions before work, every citation flagged. Your values come with it.

2

Pressure-test it

Makes the AI switch hats and attack its own answer.

Hand the analysis to the counterparty's licensing lead, who knows exactly which of these asks costs them money and which ones only sound like they do. Work through my asks as the lead, not as me. Which three of my asks do they refuse outright and why? Which two do they trade, and what do they want for them? And which one exposes a misread of the deal economics on my side. Rewrite that one correctly, then give me the compromise language for the two they will actually trade.
3

Go deeper

Pushes the work further once the basics are right.

The business lead will want to know what each clause is worth before agreeing to fight for it. Build the one-page economics model on these terms: upfront, probability-weighted milestones, and royalties run at my ask against my fallback across the licensee's forecast, with the resulting value swing per clause. Rank the clauses by dollars moved so I can see which fight is worth the relationship cost.

Before you run it

What to gather first

  • The full grant, field, royalty, diligence, improvements, sublicense, audit and termination articles
  • Which side you represent and what the alternative is if this deal dies
  • The licensee's revenue forecast and the year first sales are expected
  • Any encumbrance under the IP: federal funding, sponsored research, existing non-exclusives, reserved university rights
  • What has already been agreed in the term sheet and is off the table

Watch for

  • The net-sales definition can swing royalty value by twenty percent or more. Review deductions, rebates, chargebacks and combination-product allocation before you argue about the rate.
  • The model does not know current market royalty ranges for your sector and stage, and will produce plausible-sounding numbers if you let it. Treat every benchmark as unverified until you check it against real comparables.
  • Sublicensee survival on licensor-side termination is the recurring trap. Confirm whether sublicensees continue as direct licensees or die with the head license.
  • Encumbrances under the licensed IP (federal funding, sponsored research, existing non-exclusives) can make a promised exclusive undeliverable. The license text will not tell you they exist.
  • Improvements and derivative definitions can quietly move core IP across the table. Read those definitions against the grant clause, not on their own.

What comes back

A three-sentence bottom line naming the top three issues, the push order, and the deal-breakers; then one section per clause containing quoted operative language with section number, analysis from the client's side, paste-ready redline, and an acceptable fallback; an economics table comparing as-drafted terms, believed-typical ranges marked verified or not, the ask and the fallback; a negotiation sequence with a reason per position; and walk-away signals.

See an example of what you’ll get
*(After you answer the four questions: Northshore forecasts $180M peak sales starting 2031, Helix has two other interested licensees, only the upfront is agreed, and the vector was developed partly under an NIH grant.)* Bottom line. Three issues from Helix's side: the field is broad enough to swallow the hematologic license Helix has already promised elsewhere; the sublicense clause gives Northshore consent-free rights and zero revenue share; and the royalty is a flat 4% with no minimums against a $180M forecast. Push field first while Helix still has two alternatives, then sublicense economics, then the escalator. Walk away on field breadth and sublicense share. Field of use *(lead clause)* - *As drafted, § 1.12:* "'Field' means therapeutic uses in oncology." - *Analysis, licensor side:* Oncology covers solid tumors and hematologic malignancies. Helix's business plan reserves hematologic for a separate license to Riverpoint. As drafted, this grant forecloses that deal and Helix has no reservation clause to fall back on. - *Redline:* "'Field' means therapeutic uses in solid tumors. All other indications, including hematologic malignancies, are expressly reserved to Licensor." - *Fallback:* Solid tumors plus a 24-month right of first negotiation on hematologic, exercisable on payment of a $1.5M option fee. Royalty base - *As drafted, § 1.19:* "'Net Sales' means gross sales less reasonable deductions." - *Analysis:* "Reasonable" is an open list: the licensee defines it annually and Helix audits after the fact. Combination products are unaddressed entirely, which matters because the vector will likely be sold with a companion agent. - *Redline:* Closed list: actual returns and recalls; sales taxes and duties; separately stated freight and insurance; trade and quantity discounts actually allowed; chargebacks and rebates to government payors. Combination products allocated as A/(A+B) using average separate list prices in the same country, with Licensed Product never allocated below 50% of combination Net Sales. - *Fallback:* Same closed list, combination floor at 40%. Diligence - *As drafted, § 4.1:* "Licensee will use commercially reasonable efforts to develop and commercialize." - *Analysis:* Unenforceable as written. With exclusivity granted and no milestone, Northshore can hold the asset while it develops its internal program. - *Redline:* IND filing within 18 months; first patient dosed within 36; pivotal start within 60; BLA within 84. Any miss, subject to a 6-month cure, lets Licensor convert the grant to non-exclusive or terminate. - *Fallback:* Same milestones, 9-month cure, conversion only. Sublicensing - *As drafted, § 2.4:* "Licensee may sublicense without Licensor's consent and may retain all sublicense income." - *Analysis:* Both halves are wrong for Helix. No consent means Helix could end up in business with a party it rejected; zero share hands away the upside on the most likely exit path for this asset. - *Redline:* Consent not to be unreasonably withheld; sublicense income share of 30% pre-IND, 25% pre-pivotal, 20% thereafter; sublicenses consistent with this Agreement and terminable with it, subject to Licensor's election to recognize a compliant sublicensee directly. - *Fallback:* Notice rather than consent, with 25/20/15 splits and the direct-recognition mechanic intact. Encumbrance flag. If any of the vector work was NIH-funded, Bayh-Dole obligations attach: government license, US manufacturing preference, and march-in exposure. None of that appears in the draft. Helix cannot grant a clean worldwide exclusive until the funding history is confirmed. [UNVERIFIED - pull the grant records and any institutional patent agreement before the next call.] Economics. | Component | As drafted | Typical range | Helix ask | Fallback | |---|---|---|---|---| | Upfront | $4M | [UNVERIFIED - confirm comparables] | agreed, no change | n/a | | Royalty | 4% flat | [UNVERIFIED] | 5 / 7 / 9 tiered at $250M and $1B | 5 / 6 / 7.5 | | Minimums | none | [UNVERIFIED] | $5M from year 2 of sales | $3M | | Sublicense share | 0% | [UNVERIFIED] | 30 / 25 / 20 | 25 / 20 / 15 | Sequence. Field, then sublicense consent and share, then the royalty escalator and minimums, then the net-sales list. Field goes first because Helix's leverage is highest while two other licensees are live and the cost to Northshore of narrowing is lowest today. Walk-away. Oncology-wide field with no reservation. Consent-free sublicensing with no revenue share. Any structure where Northshore can hold exclusivity with no milestone attached. What your answers changed. The NIH answer and the two-live-licensees answer moved the most. Partial NIH funding is the entire Encumbrance flag. Without it there is no Bayh-Dole section, and I would have let Helix offer a clean worldwide exclusive it may not be able to grant. That two other licensees are live is why the push order opens on field rather than sublicense economics; if Northshore were the only party at the table, field goes last and becomes the thing I trade for the 30/25/20 sublicense split, because narrowing a field is the ask a licensee resists hardest once it knows it is the only bidder. The $180M forecast set the minimums and nothing else: $5M from year 2 of sales, $3M fallback, both priced off that number; at a $15M peak those figures are unaskable and the minimums row stays at none. Your "only the upfront is agreed" answer changed nothing: no ask on this list touches the $4M, so it confirmed what was in scope without moving a single number.
Why this prompt is built the way it is
## Framework 1. **Grant plus field is the economic perimeter.** Clarity beats breadth. "Therapeutic use in solid tumors" is a boundary; "biomedical applications" is a future dispute. 2. **Exclusivity is layered.** Exclusive even of the licensor, sole, or non-exclusive; within a field or across fields; geography; rights of first negotiation on adjacent fields. 3. **Base before rate.** The net-sales definition moves more money than the percentage. Open-ended deductions and unallocated combination products are where value leaks. 4. **Diligence with teeth or the exclusive is a shelf.** Each milestone needs an event, a date, and a consequence. 5. **Improvements come in four buckets**: licensor, licensee, joint, independent. Watch for a "derivative" definition that swallows the core IP. 6. **Sublicensing is an economics clause, not an administrative one.** Consent or notice, revenue share by stage, pass-through obligations, and what happens to sublicensees on termination. 7. **Audit rights need a threshold.** Frequency, scope, records retention, and who pays when the shortfall crosses the line. 8. **Termination is only half the clause.** Survival is the other half: sublicensees, accrued payments, confidentiality, grant-backs. 9. **Every criticism ships with a number.** Not "too narrow": the replacement language and the fallback you would sign.