zzyyxx332211

Joined 3 years ago

Karma
16
Submissions
5
Member since
Jan 24
Riccardo Bartoli · github.com/rbartoli
View on Hacker News

Recent submissions

Comment · 9 days ago
Thanks! The percentage rules don't depend on the plan as it reads the 5-hour and weekly percentages Claude Code reports, so asking at 80% and refusing at 95% works the same on Pro or Max 20x. Same for the burn rule (20 points of the 5-hour window in 10 minutes).

The ones that depend on the plan are the absolute numbers: 4 agents running at once, 12 agent starts per 10 minutes, 10M subagent tokens per 10 minutes, and the 500k context before it asks to compact. I tuned them on my own Max 5x, so on Pro they are probably too loose. You can change all of them with env vars (AGENT_GUARD_*), but presets per plan is a good idea, I'll look into it.

Comment · 9 days ago
You are correct, the first turn costs about the same either way, the difference comes after. Without compacting, every later turn re-reads the non-compacted context, and after compacting it reads a short summary.

So it only pays off if the session goes on. In my own dogfooding tests, I compacted 47 times out of 50 it had suggested.

Comment · 9 days ago
Comment · 3 years ago
I use this with Resilio Sync and custom dns names from Tailscale Magic DNS - life saver!

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton