I've been paying attention to my own workflow lately and noticed a lot of my time goes into moving stuff between AI sessions, not the actual thinking. Like I'll get an output in one session and then manually bring the relevant pieces into another so it has what it needs.
What I can't tell is how much of that is necessary vs. me just being sloppy. So I'm curious how others handle it:
- When you move from one session to another, what do you actually carry over? Just the output, or also the reasoning, the decisions, the constraints, what to avoid?
- Have you ever handed off too little and the second session went sideways? Or too much and it got lost in the noise?
- Does anyone have a mental rule for what's "enough context" to pass along?
Trying to figure out if there's a clean pattern here or if it's just inherently messy. Curious what people have landed on.
[link] [comments]