mnemon-dev/mnemon 预览 preview

mnemon-dev/mnemon

LLM监督的AI代理持久记忆——基于图谱的回忆、跨会话知识、单一二进制。兼容DeepSeek Harness、Claude Code、OpenClaw及任何代理运行时。

Project Overview项目介绍

Mnemon is an independent CLI tool that provides LLM-supervised persistent cross-session memory for AI agents. It features a four-graph knowledge store with intent-aware recall, importance decay, and automatic deduplication, running as a single local binary that requires no extra API keys. It supports multiple AI agent platforms including DSH and Claude Code, and can be installed globally via npm with a single command. After installation, mnemon setup completes the initial configuration for integration with your chosen agent platform.

Mnemon is built for developers who want to solve the common problem of AI agents forgetting critical context across sessions. It addresses issues where context compaction drops key details and long conversations push early information out of the context window. AI agents use three core primitives: remember to store new insights, link to connect related knowledge, and recall to pull relevant historical context into the current session. This persistent memory accumulates value over time, steadily improving an agent’s performance on long-running development projects and eliminates repeated re-explanation of context.

Mnemon is released under the open-source Apache 2.0 license. It requires Go 1.24 or newer to build from source, and existing users can upgrade any npm-installed version with the mnemon update command at any time. All memory data is stored locally on the user’s device, and the tool requires no extra cost beyond an existing LLM subscription, with no additional API keys needed for core functionality. First-time users get an interactive setup process after installation to configure their agent integrations.

Mnemon 是一个为 AI 代理提供大语言模型监督的持久化跨会话记忆的独立命令行工具。它采用四图知识存储架构,具备意图感知召回、重要性衰减和自动去重功能,以单一本地二进制文件运行,无需额外 API 密钥,仅需一条命令即可完成初始设置。它同时支持 DSH 和 Claude Code 等多个 AI 代理平台,可通过原生清单集成到 DSH 生态中。

这款工具面向需要长期积累项目知识、解决 AI 代理跨会话遗忘问题的开发者。日常使用中,AI 代理可以通过 remember、link、recall 三个原生命令来存储、关联和召回知识,在新会话中自动拉取历史相关信息,避免关键决策和上下文因上下文窗口压缩丢失。它让开发者可以持续积累项目上下文,提升 AI 代理在长期开发项目中的表现。

它使用 Apache 2.0 开源协议,要求 Go 1.24+ 版本进行构建,支持通过 npm 在 macOS、Linux 和 Windows 系统全局安装,可随时使用 mnemon update 命令升级。工具完全运行在本地,所有数据存储在用户本地设备,不依赖额外的第三方云服务,仅需用户现有大语言模型订阅即可使用,无需额外付费。

Pre-install check安装前体检Compatibility · Security兼容性 · 安全性 no risk signal found未发现风险信号
  • This site's static screen found no obvious risk signal (stars, license, activity, manifest)本站静态筛查没发现明显风险信号(星标、许可证、更新活跃度、清单完整度)
DSH walks through these 9 checksDSH 会逐条核对这 9 项

Compatibility兼容性

  • DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
  • External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时(Electron / Python / Docker 等)是否齐备
  • Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突:命令名、skill / tool 重名、端口占用、重复 MCP 注册

Security安全性

  • Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致,是否归档或长期停更
  • Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
  • curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
  • Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
  • Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需(filesystem / network / shell / clipboard)
  • Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限,以及卸载与回滚方式

Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照,不能替代安全审计。

Or use CLI install (for developers)或使用命令行安装(适合开发者)

CLI Install命令行安装

dsh plugin --profile web add dsh-mnemon

把 mnemon-dev/mnemon 加入你的 DSH 配置(web profile)即可启用。

READMEREADME

Mnemon Logo

Mnemon

English · 中文

npm version GitHub release GitHub stars Go 1.24+ CI License: Apache-2.0

LLM-supervised persistent memory for AI agents.


LLM agents forget everything between sessions. Context compaction drops critical decisions, cross-session knowledge vanishes, and long conversations push early information out of the window.

Mnemon gives your agent persistent, cross-session memory — a four-graph knowledge store with intent-aware recall, importance decay, and automatic deduplication. The mnemon memory path remains one local binary with zero API keys and one setup command.

Mnemon ships one executable with two separate surfaces. Memory stays at the mnemon root; Agency Preview lives at mnemon agency ... and adds durable, project-local responsibility and effect admission to an existing Pi agent. Agency does not replace Memory or the Agent Runtime.

Claude Max / Pro subscriber? Mnemon works entirely through your existing subscription — no separate API key required. Your LLM subscription is the intelligence layer. Two commands and you're done.

Why Mnemon?

Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要,完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →

← 上一个 Prev superdesign-skill 下一个 Next dsh-mobile-apk →