mosaic-memory-compress
为 DeepSeek Harness 提供基于自然遗忘曲线的就地对话压缩：近期对话逐字保留、中期结构化脱水（零 LLM 成本）、久远对话折叠为有界摘要对——有限记忆预算上的无限对话。
Forgetting-curve in-place dialogue compression for DeepSeek Harness: recent turns stay verbatim, middle turns are structurally dehydrated (zero-LLM), ancient turns fold into a bounded summary pair — unbounded conversations on a bounded memory budget.
怎么安装
dsh plugin add mosaic-memory-compress 简介
MosaicMemoryCompress **A generic, pluggable stateless dialogue compression algorithm** — works with any LLM agent framework, and ships a ready-to-use adapter module for **DeepSeek Harness (DSH)**. LLM conversations grow linearly. MosaicMemoryCompress keeps them bounded — automatically, invisibly, and without the user ever knowing what a "Session" is. How It Works Your message array (R rounds, oldest → newest): Round 1 ────→ Round (R-30) │ Heavy zone → ALL → 2 msgs Round (R-29) → Round (R-10) │ Light zone → structural truncation, count unchanged Round (R-9) ────→ Round R │ Raw zone → keep as-is…
推荐参考
信息
- 协议
- MIT
- 语言
- TypeScript
- GitHub 星标
- 3
- 月下载
- –
- 最近更新
- 2026-09-12
- 创建于
- 2026-06-08
基础安全检查
- 检查结果
- 无
- 收录来源
- curated:awesome-dsh-plugin.com, curated:awesome-dsh-plugin/awesome-dsh-plugin
- 主题标签
- agent-memory, ai-agents, context-compression, context-management, context-window, conversation, deepseek, deepseek-harness, deepseek-harness-plugin, developer-tools, dialogue-compression, dsh-memory, dsh-plugin, forgetting-curve, llm, memory-management, session-management, token-efficiency, typescript
同类推荐
hindsight
vectorize-io/hindsight/tree/main/hindsight-integrations/coding-agents
Hindsight：会学习的 Agent 长期记忆系统，自动召回/保存、知识页、深度反思与按仓库隔离的记忆银行。
firecrawl-mcp-server
firecrawl/firecrawl-mcp-server
🔥 官方Firecrawl MCP服务器 - 为Cursor、Claude及任何其他LLM客户端添加强大的网页抓取和搜索功能。
dsh-mnemon
omdsh-dev/dsh-mnemon
由 Mnemon 驱动的 DeepSeek Harness（DSH）跨 Agent、本地优先的持久记忆插件。它可在支持 Mnemon 的 Agent 之间共享长期记忆，并提供运行时记忆、可检索项目档案、语义召回、知识图谱和 Sidebar UI。
get-shit-done
gsd-build/get-shit-done
由 TÂCHES 开发的轻量且强大的元提示、上下文工程和规格驱动开发系统，适用于 Claude Code。