Richard Francis

An agent is fixing a bug and starts touching files outside the requested scope. Before switching tools, I'd check whether the task has a reproducible failure and a clear boundary. Without those, the next agent gets the same ambiguity.

If the task is well defined, I'd give Codex and Claude Code the same starting commit and stop at the first reviewable diff. I'd compare unrelated edits, tests actually run, and the human work needed before merge. A passing test alone misses that cost.

If you've switched mid-task, what triggered it? Did the second tool make the patch easier to review, or did you mostly improve the brief on the second try?

My LinkedIn inbox is half spam these days. Pitches right after connecting, fake "quick question" messages, AI written compliments about my profile.

But I still reply to some strangers. Usually when they mention something real I posted or just ask a normal human question without selling right away.

I'm curious where your line is?

What makes you reply to a DM from someone you don't know? And what makes you ignore it instantly?

I build my app alone after my day job, with no ad budget. So far the only thing that really worked for me is writing short, personal posts on Threads. One story about my family got over 300k views and brought more signups than everything else combined. Posts that explain features barely get noticed.

But I can't tell if that was luck or something I can repeat.