context-compression

LLM-free lossless* compaction engine for DeepSeek Harness

Auxiliary models for DeepSeek Harness: vision understanding and context compression through dedicated model routes. DeepSeek Harness 辅助模型插件：为视觉理解、上下文压缩、审批审查、子代理、会话标题与图片生成提供独立的模型路由、工具与系统提示，全程不触碰主对话模型。

Guarded context compaction for DeepSeek Harness (dsh): the LLM proposes, deterministic guards dispose — eager per-atom shrink (extract/summary/false under verbatim guards) + lazy reference-graph eviction (0-LLM) + byte-exact recall from an append-only log. 守卫式上下文压缩：LLM 只提议、确定性守卫裁决——逐原子缩放 + 惰性引用图剪枝（0-LLM）+ 追加式日志逐字节召回；压缩率精确兑现，历史永不销毁。

Headroom context-compression proxy integration for DeepSeek Harness: detect/install/start Headroom, auto-configure the DeepSeek compressed route, one-click route-switch UI. Built on headroomlabs-ai/headroom (Apache-2.0).

Context Assembler DSH — DeepSeek Harness 上下文汇编插件：按话题块编排上下文，最大化每 token 互信息密度

Manual budget-targeted context compression for DeepSeek Harness: the conversation model picks ranges, a cheap flash-tier route writes the summaries.

DSH plugin for agent-driven span compaction: compress chosen conversation spans into self-written checkpoints instead of the official head-anchored full-context sweep.

人脑式上下文压缩逻辑，大幅减少长任务上下文占用以及token消耗

该仓库暂未提供项目说明。

A context engine for DeepSeek Harness — reversible, token-budgeted compression of the live context window.

A customized natural-language compaction plugin. It gives you the ability to precisely trim context — sometimes even to fix a malfunctioning context.

⚡ Context economy layer for DSH: symbol/import graph indexing with 80-93% token savings

DSH 插件 · 会话级压缩（官方 auto 只能全局） / Per-session compaction; official auto is global-only