Rudekwydra/burnless
View on GitHub ↗ ★ 7 Python
Stop replaying the transcript: capsule session state + rolling memory for LLM CLIs (Claude Code, Codex). Measured: −90.3% vs no-cache / −30% vs cached replay at turn 10, reproducible bench. MIT, provider-agnostic.
3 mentions on Hacker News by 1 person: 3 stories, 0 comments · first 2026-05-04 · latest 2026-07-30
20082026
-
▲ 1
Show HN: Burnless makes LLM context a protocol. Token savings are a side effectstory by rudekwydra · 2026-07-30 · 1 comment
-
▲ 2
Show HN: Burnless – 1,590 tokens of context for a 1.44M-token workdaystory by rudekwydra · 2026-07-29 · 1 comment
-
▲ 2
Burnless – open protocol that cut my API bill 16x on my heaviest dev daystory by rudekwydra · 2026-05-04 · 0 comments