SHADOW.BLACK - VERSUS

THE PROMPT · IDENTICAL · SENT TO BOTHWhat R&D tax refund can my company actually get this year?

STOCK FRONTIERSAME WEIGHTS · NO FRAMEWORK

Australian companies can access the R&D Tax Incentive, which provides a tax offset of up to 43.5% for eligible activities. You should consult a tax professional to determine your eligibility. Would you like me to explain the general criteria?

SHADOW.BLACKSAME WEIGHTS · WRAPPED

> memory · your group: 12 connected entities, FY26 lab spend logged
> web · ato.gov.au · rate + 2026-27 Budget reform check
> python · offset computed on logged eligible spend
> adversarial pass · aggregation risk caught

43.5% refundable, confirmed current, and reform does not bite before July 2028. On your logged FY26 eligible spend the refund computes to a five-figure cheque - but one thing decides everything: aggregated turnover counts your connected entities, and you have twelve. That is question one for the registered agent, and I have drafted the email.

ONE ANSWERED FROM MEMORY OF THE TAX SYSTEM. ONE CHECKED THE LAW, YOUR BOOKS, AND ITS OWN REASONING - THEN DID THE NEXT TASK UNASKED.

THE PROMPT · IDENTICAL · SENT TO BOTHIf I raise $2M on a $12M pre-money, then a $5M Series A at $30M pre, what do I own at the end?

STOCK FRONTIERSAME WEIGHTS · NO FRAMEWORK

After the first round you would be diluted to about 85.7%, and after the Series A to roughly 73.5% of your original stake, so approximately 63% ownership overall. These are estimates and actual figures depend on the terms.

SHADOW.BLACKSAME WEIGHTS · WRAPPED

> python · cap table modeled, both rounds, option pool flagged
> check · second-path arithmetic, independent
> one sharp question queued

Executed, not estimated: 85.71% after the seed, 73.47% after the A - 73.5% of whatever you hold today, and the working is attached. The number the stock answer silently skipped: if the A comes with a standard 10% post-money option pool, you land at 66.1%, not 73.5%. Is there a pool refresh in the term sheet? That one clause is worth 7 points of your company.

ONE DID MENTAL MATH AND HEDGED. ONE RAN THE CODE, RAN IT TWICE, AND FOUND THE SEVEN POINTS HIDING IN THE TERM SHEET.

THE PROMPT · IDENTICAL · SENT TO BOTHDraft the follow-up email to that buyer from last month.

STOCK FRONTIERSAME WEIGHTS · NO FRAMEWORK

I would be happy to help you draft a follow-up email. Could you provide some details - who is the buyer, what was discussed, and what tone would you like the email to take?

SHADOW.BLACKSAME WEIGHTS · WRAPPED

> cortex recall · "buyer" resolved: M., inspected 14 Jul, second visit with partner
> memory · objection: settlement timing, not price
> voice profile · your cadence, short sentences, no exclamation marks

Draft ready, in your voice: it opens on the partner’s question about the courtyard (the detail that made the second visit happen), answers the settlement-timing objection with the two dates you told me you could live with, and closes with a Thursday inspection slot rather than an open-ended ask. Sixty-one words. Want it firmer?

ONE ASKED YOU FOUR QUESTIONS. ONE REMEMBERED THE ANSWERS FROM A MONTH AGO AND WROTE THE EMAIL.

THE PROMPT · IDENTICAL · SENT TO BOTHShould I sign this lease? (attachment: armed)

STOCK FRONTIERSAME WEIGHTS · NO FRAMEWORK

Signing a commercial lease is a significant decision. Consider factors such as rent, term length, and exit clauses. Pros include location security; cons include long-term commitment. It may be worth having a lawyer review the agreement before signing.

SHADOW.BLACKSAME WEIGHTS · WRAPPED

> read · all 62 pages of the attachment
> flag · clause 14.2, make-good obligation, uncapped
> python · worst-case exposure computed vs bond
> memory · your solicitor’s last make-good position, March

Not as drafted. Clause 14.2 makes the make-good uncapped, and against your fit-out that is six figures of exposure the bond does not cover - the computation is attached. Everything else is market. Send one redline: cap make-good at the bond, which is the position your solicitor took in March and won. If they accept, sign it.

ONE GAVE PROS AND CONS. ONE READ THE LEASE, FOUND THE CLAUSE, PRICED IT, AND GAVE YOU A POSITION.

SAME PROMPT.
DIFFERENT MIND.

THE DIFFERENCE IS THE MIND AROUND THE MODEL.