The four heights
The same task, four distances: today's deadline, the next reviewer, the stuck moment, the pattern.
Execute — do the immediate task
+I need to send my investment recommendation memo to Miguel in Corporate Strategy and then to Tanya…
Execute — do the immediate task
+I need to send my investment recommendation memo to Miguel in Corporate Strategy and then to Tanya in Treasury for approval, signers in that order, with a decision by next Friday. Before that, confirm the supporting cash-flow table matches the forecast model and that the IRR, payback and sensitivity tabs are intact.
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Improve — make it easier to accept
+Before I send this to Miguel in Corporate Strategy, make the recommendation easy to act on: put the…
Improve — make it easier to accept
+Before I send this to Miguel in Corporate Strategy, make the recommendation easy to act on: put the headline expected return and required capital on the first page, surface the worst-case cash-flow number, tag the assumptions that would make Tanya refuse funding, and make the sensitivity chart clickable so a reviewer can find it instantly.
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Decide — diagnose the stuck moment
+The forecast model produced an IRR three percentage points higher after I consolidated the June…
Decide — diagnose the stuck moment
+I ran the model and the IRR jumped three points unexpectedly
The forecast model produced an IRR three percentage points higher after I consolidated the June entries; Miguel needs a clean yes/no by Friday and Tanya will kill it if the cash cushion under worst-case is below eight months. I don't know if the change came from a hidden formula or an input error. What’s the most likely source and the fastest way to prove it before I circulate?
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Become — change the pattern
+Over the last six recommendations I keep finding last-minute breaks in linking and wrong…
Become — change the pattern
+I keep fixing similar model errors at the last minute
Over the last six recommendations I keep finding last-minute breaks in linking and wrong assumptions that cost me nights before sign-off. I lose credibility and delay approvals. What one habit should I adopt to stop these recurring model-logic and linking errors?
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Where the evidence lives
Who was seen doing this, and what people really ask.
Commercial AnalystCrm AnalystData Governance AnalystExperimentation AnalystFinancial AnalystFinancial Reporting Analystalso: Recommend investments and timing
Software tasks in the LLOS Work Atlas come from evidence, never a feature list: careers attested to do the work, real job descriptions, and the questions people actually ask (with their view counts). Facets — feature, workflow, troubleshoot, administer, deploy, scale — are open metadata: the work decides, not a taxonomy.
Copyright © LLOS.ai · 2026 — original pedagogy, voice, and design — all rights reserved.
The rest of the map
Same library, five ways in.