do you trust auto-approve mode for terminal commands or still check every single one?
by•
clicking approve every ten seconds while an agent installs packages or runs local test scripts kills the whole flow of fast prototyping.
at the same time, turning on full auto-run makes me nervous that one hallucinated shell command might wipe local environment files or drop a dev database.
how much terminal freedom do you give your coding agents before stepping in?
14 views
p/many-displays
p/runevr
p/producthunt
p/google-cloud-platform
p/reso
p/kodlo
p/ai-context-flow
p/htmldrag
p/dex-by-exmergo
p/jot-2-0
p/beamfy
p/refs
p/dino-kids
p/nexupira
p/duckfightclub
p/nudgeforme
p/seshka
p/talanoa-2
p/murror
p/supabase
Replies
I run in auto mode, but with fences. Right now I have 89 allow rules and 25 deny rules, plus 94 hooks that check commands before they run. Dropping or truncating tables, recursive deletes and force pushes are blocked outright, dev or production, and anything that can't be undone, like deploys or real emails, still waits for my OK.