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 have the new draft policy impact table and the historical dataset of affected transactions. Send…
Execute — do the immediate task
+I have the new draft policy impact table and the historical dataset of affected transactions. Send the impact workbook and a two-paragraph executive summary to Lina in Policy and to Omar in Risk, asking for their signoff by Tuesday and noting which sheets you validated. Confirm the scenario formulas update and the sensitivity rows are correct before sending.
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 present this to the regulator group, make the policy impacts easy to challenge: put…
Improve — make it easier to accept
+Before I present this to the regulator group, make the policy impacts easy to challenge: put headline monetized impact and affected population at the top, highlight the three largest drivers, and flag assumptions that would make the impact swing by more than 10 percent. Make the counterfactual clear and attach the two alternate sensitivity runs.
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
+I must reply to Ana in Policy with a single number today; the ranges are wide and Risk needs more…
Decide — diagnose the stuck moment
+Regulators asked for an immediate point estimate and I only have ranges.
I must reply to Ana in Policy with a single number today; the ranges are wide and Risk needs more time. What single estimate is defensible to give now, how do I qualify it in one sentence so the regulator understands uncertainty, and what follow-up data request should I open to shrink the range?
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 past year, policy changes cost us six rushed reruns and two escalations to legal. I want a…
Become — change the pattern
+Every time a policy changes we scramble to rerun models and miss deadlines.
Over the past year, policy changes cost us six rushed reruns and two escalations to legal. I want a repeatable pattern to reduce emergency work: should we keep a standing 'policy change' prototype workbook with parameterized inputs, require a policy impact checklist before modeling, or institute a policy change triage meeting? Recommend one habit change, steps to implement in two weeks, and how to measure if it actually saves time and reduces escalations.
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.
Development Policy SpecialistMacroeconomistMicroeconomistRevenue AnalystTrade EconomistPolitical Scientistalso: Evaluate the impact of policies and laws
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.