Set auto-compact window to 400k to cut token usage on 1M-context models
mariofilhoml · x · 2026-11-12
A practical tip circulating on X: when using Claude Code or similar harnesses, set the auto-compact window to 400k and usage drops significantly. Another user adds that this works for any 1M-context model in any harness — preventing the model from always filling context to the million-token limit before compaction kicks in.
More from coding & agent
- Game dev workflow: AI-generated sprite sheets turned into character walk GIFs — algo_diver · 2026-11-12
- Agent reads email threads and pre-books your apartment viewing on Gcal automatically — braelyn_ai · 2026-11-12
- AntiFox AI Pentesting Platform Linked to Attacks That Stole AWS Production Secrets — Thionne_WTZ · 2026-11-12
- Dev builds npx driftcost CLI to check if Claude Code token usage silently shifted between Opus versions — Tough_Stay8909 · 2026-11-12
- AFK Pilot ships standalone desktop app to run coding agents without an IDE — PawelHuryn · 2026-11-12
- Gemini 4 Carbon now usable in VS Code and Cursor via AFK Pilot extension — PawelHuryn · 2026-11-12