Auto-compaction, auto-continue, auto-planning. Replaces OpenCode's default auto-compact/auto-continue lifecycle with one that creates a portable, persistent PLAN.md document. One persistent source of truth for your project that grows as the agent works autonomously. Continue using it in the current session or start a new session with it.
为 DeepSeek Harness 极简模式增加自动上下文压缩、/compact、/context 和模型主动压缩，解决极简模式长任务无法持续工作的问题。
Windows skill for ChatGPT Work Local & Codex Desktop/Local: configure context window token limits, effective percentages and auto-compaction thresholds. Global/per-model settings, preview, backup and restore. No Python setup. 本地上下文窗口调整技能：自定义Token上限、有效比例、自动压缩阈值，支持全部或指定模型、预览、备份恢复；无需安装Python。
Defense in depth for Claude Code context. Multi-layer preservation system: hooks + skills + agent + rules. Pure bash, zero dependencies.
70-99% context compression for OpenCode v2: ContextPack modules, per-request budgeting, eager tool-result paging with byte-exact recall, a per-project library and warm carry-in between sessions.
To associate your repository with the auto-compaction topic, visit your repo's landing page and select "manage topics."