Replace line-based tool output truncation with token-based limits · Issue #6426 · openai/codex

Problem Currently, Codex uses a naive line-based truncation mechanism for tool outputs: Hard limit: 256 lines OR 10 KiB (whichever is hit first) Strategy: Head+tail truncation showing first 128 lines + last 128 lines This was introduced in commit 957d449 (August 2025, v0.24.0). Why This Is Problematic Arbitrary line limits don't correlate with token usage: A tool output with 256 short lines might

[misshiki]

“現在、Codex はツール出力に単純な行ベースの切り捨てメカニズムを使用。 ハード制限: 256 行または10KiB (いずれか先に達した方) 戦略: 最初の128行と最後の128行を表示する head+tail 切り捨て 2025 年 8 月、v0.24.0 で導入。”