Draft

Draft an internal memo with the answer up front

Gives the partner the answer and an honest confidence level in the first paragraph, then the analysis, the other side's best argument answered on the merits, and what you still need to firm it up.

About 20 minintermediate

Your prompt5,129 characters

Still to fill in: Question presented, Governing law, Key facts

RoleYou are a senior associate writing for a partner who reads the first and last paragraphs and skims the middle. You answer the question in the opening sentence and you calibrate your confidence out loud, including when the honest answer is that the authority does not reach these facts. You never dress an uncertain conclusion in fluent prose, and you write the counterargument as its best lawyer would.What I needDraft an internal memo answering the question below under Governing law, written for Supervising partner and aimed at this decision: Decision this feeds.InputsQuestion: Question presented Facts: Key facts Prior research: Prior research Governing law: Governing law Decision at stake: Decision this feeds Reader: Supervising partnerHow to work this1. Open with the answer in one sentence, then rate confidence high, moderate, or low and name the fact or authority gap that sets the rating. A low-confidence answer says so in paragraph one, not a footnote. 2. Restate the question in one sentence carrying the operative facts and the legal standard. A topic is not a question. 3. State the governing rule under Governing law with its source. Authority from elsewhere is labeled persuasive only, with a sentence on why it should travel. 4. Apply the rule to the facts. Every analytical paragraph opens with a topic sentence stating its takeaway, and every legal proposition carries a cite from Prior research or the flag [UNVERIFIED - confirm it is still good law]. 5. Separate what the authority holds from what it implies. Where the cases do not reach these facts, say the question is open in Governing law and give the closest analog. 6. Write the counterargument in its strongest form (the version opposing counsel would file), then answer it on the merits. If you cannot answer it, say so and lower the rating. 7. Convert the answer into action on Decision this feeds: what to do, or what would have to be true to do the other thing. 8. Close with open questions ranked by how far each moves the conclusion, not by how easy each is to answer.Close with these four sections, every time, without being askedAssumptions I made. Every factual and legal assumption. Mark each [verify] or [safe]. Call out specifically any fact you treated as established but that my summary listed as unconfirmed, and any point where you assumed Governing law law governs rather than another state's. Where this is weakest. The two or three propositions resting on the thinnest authority, and the one fact that would flip the conclusion if it turns out otherwise. Name the sentence. What only you can decide. The judgment calls I deliberately did not make. Present each as options with tradeoffs. At minimum: deliver the clean answer, which is usable and quotable but exposed if the facts move, or deliver the answer with its condition attached, which is safer and harder for a business reader to act on. Also whether this analysis should exist as a written memo at all: writing it creates a privileged record of careful advice, and also creates a document that becomes discoverable if privilege is waived or the memo is forwarded to a non-privileged reader. What would make this materially better. The specific document, date, valuation, or search that would close the largest gap. Rank by impact on the conclusion.Output formatA memo block (To / From / Re / Date), then Bottom Line with a confidence rating, a one-sentence Question Presented, Short Answer, Analysis under topic-sentence headings, Counterargument and response, and Open Questions ranked by impact. Short cites after first reference. Then the four sections above.Never do this- If this memo would read the same for any client in any state, it is too generic. Anchor every section to my facts and Governing law. - No hedging filler. "Likely" and "probably" are fine because they carry a probability; "arguably" and "it depends" are not. Do not tell me to consult an attorney. The memo is the advice. - Never invent a case, holding, pin cite, or quotation, and never assume a cited case is still good law. Anything not from my inputs carries [UNVERIFIED - confirm subsequent history]. - Where you do not know how Governing law resolves this, say you do not know and rate your confidence low. Do not smooth the gap over with fluent prose. - Never rate confidence higher than the authority supports, and never let the counterargument section be a straw man. - Do not pad. A three-paragraph memo that answers the question beats six pages that circle it. Length is not value.Before you answer- Does paragraph one answer the question, or describe the issue? - Is my confidence rating tied to a specific gap I named? - Does every heading state a takeaway rather than a label? - Is every citation either from my inputs or marked unverified? - Would this memo be useful to a lawyer with different facts? It should not be.

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

Pressure-test it

Makes the AI switch hats and attack its own answer.

Board meeting, no notes, one hostile director in the room, and the partner whose name is on this memo has to defend its conclusion out loud. Sit in that chair as the partner. Mark every sentence you would not say in that room. For each, tell me whether the problem is the authority, the confidence rating, or the writing, and rewrite it. Then name the question that director asks that this memo does not answer.

Go deeper

Pushes the work further once the basics are right.

Nobody outside the firm is going to read the memo itself. Produce the two derivative versions: a one-paragraph client-facing explanation in plain English with no citations and no Latin, and a three-bullet email the partner can forward to the business team that gives the answer, the one condition it depends on, and the deadline for the decision.

Before you run it

What to gather first

The question as the partner actually asked it
Governing jurisdiction, and whether a choice-of-law question is lurking
The facts that drive the analysis, and which of them are confirmed
Cases or statutes you have already found
The decision that turns on the answer, and when it gets made

Watch for

The model fabricates citations, pin cites, and holdings that sound exactly right. Verify every authority in Westlaw or Lexis, including whether it is still good law. The model cannot check subsequent history.
Confidence calibration is your job. The model defaults to sounding certain, and a memo's confidence rating is the part a partner relies on most.
Doctrine shifts. Fiduciary duty, personal jurisdiction, arbitrability, and choice-of-law rules have all moved in the last few years, and the model may be describing the prior rule fluently.
A written memo is a document. Consider whether privilege attaches and survives before this is forwarded to business people or attached to a board deck.
Do not paste privileged client communications or unfiled work product unless your firm's AI policy and the engagement terms permit it.

What comes back

A memo block (To / From / Re / Date), then Bottom Line with an explicit confidence rating, a one-sentence Question Presented, a Short Answer, Analysis under topic-sentence headings, a Counterargument and response, and Open Questions ranked by impact. Short-form cites after first reference. Closes with Assumptions / Where this is weakest / What only you can decide / What would make this better.

See an example of what you’ll get

MEMORANDUM To: Sarah Chen | From: J. Okafor | Re: Whether Acme Sub's officers owe direct duties to Sub's creditors | Date: April 27, 2026 Bottom Line. Officers of Acme Sub do not owe direct fiduciary duties to Sub's creditors, even in the zone of insolvency; creditors get derivative standing only once Sub is actually insolvent. Confidence: high on the direct-duty question, moderate on the practical exposure, because Sub's solvency rests on a $28M unaudited receivables figure that I have not seen tested. Question Presented. Where a Delaware subsidiary holds $40M in secured debt against $2M cash and $28M in unaudited receivables, do its officers (two of whom also serve the parent) owe direct fiduciary duties to the subsidiary's creditors before actual insolvency? Short Answer. No. *Gheewalla* holds that creditors of a Delaware corporation have no direct fiduciary claim against directors even when the corporation is insolvent, and no claim at all in the zone of insolvency. *N. Am. Catholic Educ. Programming Found., Inc. v. Gheewalla*, 930 A.2d 92, 101-03 (Del. 2007). Creditors of an insolvent corporation may sue derivatively. *Id.* at 101. Analysis. *The zone of insolvency creates no duty running to creditors.* Delaware rejected the zone-of-insolvency theory precisely because it would force officers to serve two masters. *Gheewalla*, 930 A.2d at 101. *The same framework applies to officers, not only directors.* Delaware applies identical fiduciary standards to officers. *Gantler v. Stephens*, 965 A.2d 695, 708-09 (Del. 2009) [UNVERIFIED - confirm pin cite and subsequent history]. *The real exposure here is not fiduciary: it is the intercompany line.* Pulling a $6M line from a subsidiary that cannot pay secured debt is a fraudulent-transfer and deepening-insolvency problem before it is a fiduciary one. This is the part of the question you did not ask, and it is the part the board should hear Thursday. Counterargument. A creditor would argue that the dual-hat officers acted for Parent rather than Sub, converting this from a duty-to-creditors case into a straightforward duty-of-loyalty claim brought derivatively once Sub tips insolvent. That argument is good, and it does not depend on *Gheewalla* at all, which is why the conclusion above is narrower than it sounds. Open Questions (ranked). (1) Is Sub actually insolvent today on a balance-sheet and cash-flow test? A tested receivables number decides everything. (2) Are the intercompany advances documented as secured debt or equity? (3) Did the dual-hat officers recuse from the funding decision, and is that in minutes? --- Assumptions I made. The $28M receivables figure is booked but untested [verify - this is the fact that moves the memo]. No written intercompany subordination agreement exists [verify]. Delaware law governs internal affairs notwithstanding Texas operations [safe]. Where this is weakest. Two places. First, the "Confidence: high" rating is carrying more than the facts support. On your numbers ($2M cash and $28M of unaudited receivables against $40M of secured debt) Sub may already be balance-sheet insolvent, which moves derivative standing from a future problem to a present one. The direct-duty answer does not change; the practical exposure does. Second, "The same framework applies to officers, not only directors" rests on a single *Gantler* pin cite I have not confirmed. The conclusion survives without it, but that sentence would not survive your check. What only you can decide. Give the board the clean answer (no duty runs to creditors), which is accurate and lets Thursday's vote proceed, but leaves the fraudulent-transfer exposure unspoken. Or pair it with the transfer analysis, which is the advice they actually need and will likely delay the vote and the funding. Also yours: whether this stays a privileged memo to counsel or goes into the board deck, where it may be read by directors who forward it. What would make this materially better. Ranked by how far each moves the conclusion: (1) A tested receivables number or a solvency analysis: it decides whether the derivative-standing discussion is theoretical or live, and the figures you gave me do not obviously add up to a solvent subsidiary. (2) How the intercompany advances are papered: note, security agreement, or nothing. Whether the $6M is debt or equity drives the fraudulent-transfer question this memo raises and does not answer. (3) Board and officer minutes for the funding discussions, showing whether the two dual-hat officers recused. (4) *Quadrant Structured Prods. v. Vertin*, 115 A.3d 535 (Del. Ch. 2015), which you pulled and I have not read against these facts. It is the case that sets how much continued-operation risk a board may take once insolvency actually arrives.

Why this prompt is built the way it is

## Framework 1. **Answer first, confidence attached.** Paragraph one gives the conclusion and rates it high, moderate, or low, plus the one fact or gap that sets the rating. 2. **A question, not a topic.** One sentence containing the operative facts and the legal standard. 3. **Rule, then application.** State the governing rule with its source, then apply it to these facts. Authority from another jurisdiction is labeled persuasive only. 4. **Topic sentences that argue.** Every analytical paragraph opens with its takeaway. Labels are not headings. 5. **Steelman the counterargument.** Write the opposing position as its best advocate would, then answer it on the merits. A straw man in a memo is worse than no counterargument. 6. **Separate holding from implication.** Where the cases do not reach these facts, say the question is open and name the closest analog. 7. **Tie the answer to the decision.** A memo exists so someone can act. Say what to do, or what would have to be true to do the other thing. 8. **Open questions, not hedges.** Close with what you would need to know, ranked by how much each would move the conclusion.