The four heights
The same task, four distances: today's deadline, the next reviewer, the stuck moment, the pattern.
Execute — do the immediate task
+Find every occurrence of the feature flag name enableNewSearch across the codebase and open the…
Execute — do the immediate task
+Find every occurrence of the feature flag name enableNewSearch across the codebase and open the list so I can review each usage before I flip it for Friday’s deploy.
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 flip enableNewSearch, surface the most risky uses of that flag: show callbacks that run on…
Improve — make it easier to accept
+Before I flip enableNewSearch, surface the most risky uses of that flag: show callbacks that run on startup, any server-side checks, and configuration files; put the number of references and the top three files with the heaviest usage up front.
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 searched for enableNewSearch and got dozens of hits from documentation and comment blocks; I need…
Decide — diagnose the stuck moment
+I searched for enableNewSearch but got too many irrelevant hits
I searched for enableNewSearch and got dozens of hits from documentation and comment blocks; I need to know which are actual runtime checks versus non-executable mentions. I’m worried I’ll miss a server-side switch that controls behavior in production. How do I reliably distinguish executable references from noise and what’s the next step to ensure I don’t flip the flag prematurely?
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Become — change the pattern
+Across projects I waste 30–60 minutes per feature toggling task sifting through false positives…
Become — change the pattern
+Search results frequently include docs and comments that waste my time
Across projects I waste 30–60 minutes per feature toggling task sifting through false positives from docs and generated files. Where should I change my workflow so searches return actionable hits: naming conventions, ignore lists, or a small pre-search checklist? Recommend the single highest-impact habit to adopt first.
Pasted it? When the reply comes back, push once: ask it to sharpen the weakest part. — Did this prompt help?
Next to this one
Other code editor work people do in Visual Studio Code.
Every task here came from the work, not from a feature list — which is why the prompts name what you want done and never the button that does it. The tool changes; the work does not.
Copyright © LLOS.ai · 2026 — original pedagogy, voice, and design — all rights reserved.
The rest of the map
Same library, five ways in.