# memu - Doramagic AI Context Pack

> 定位：安装前体验与判断资产。它帮助宿主 AI 有一个好的开始，但不代表已经安装、执行或验证目标项目。

## 充分原则

- **充分原则，不是压缩原则**：AI Context Pack 应该充分到让宿主 AI 在开工前理解项目价值、能力边界、使用入口、风险和证据来源；它可以分层组织，但不以最短摘要为目标。
- **压缩策略**：只压缩噪声和重复内容，不压缩会影响判断和开工质量的上下文。

## 给宿主 AI 的使用方式

你正在读取 Doramagic 为 memu 编译的 AI Context Pack。请把它当作开工前上下文：帮助用户理解适合谁、能做什么、如何开始、哪些必须安装后验证、风险在哪里。不要声称你已经安装、运行或执行了目标项目。

## Claim 消费规则

- **事实来源**：Repo Evidence + Claim/Evidence Graph；Human Wiki 只提供显著性、术语和叙事结构。
- **事实最低状态**：`supported`
- `supported`：可以作为项目事实使用，但回答中必须引用 claim_id 和证据路径。
- `weak`：只能作为低置信度线索，必须要求用户继续核实。
- `inferred`：只能用于风险提示或待确认问题，不能包装成项目事实。
- `unverified`：不得作为事实使用，应明确说证据不足。
- `contradicted`：必须展示冲突来源，不得替用户强行选择一个版本。

## 它最适合谁

- **正在使用 Claude/Codex/Cursor/Gemini 等宿主 AI 的开发者**：README 或插件配置提到多个宿主 AI。 证据：`README.md` Claim：`clm_0002` supported 0.86

## 它能做什么

- **命令行启动或安装流程**（需要安装后验证）：项目文档中存在可执行命令，真实使用需要在本地或宿主环境中运行这些命令。 证据：`README.md` Claim：`clm_0001` supported 0.86

## 怎么开始

- `pip install memu-py` 证据：`README.md` Claim：`clm_0003` supported 0.86
- `git clone https://github.com/YOUR_USERNAME/memU.git` 证据：`README.md` Claim：`clm_0004` supported 0.86

## 继续前判断卡

- **当前建议**：需要管理员/安全审批
- **为什么**：继续前可能涉及密钥、账号、外部服务或敏感上下文，建议先经过管理员或安全审批。

### 30 秒判断

- **现在怎么做**：需要管理员/安全审批
- **最小安全下一步**：先跑 Prompt Preview；若涉及凭证或企业环境，先审批再试装
- **先别相信**：角色质量和任务匹配不能直接相信。
- **继续会触碰**：角色选择偏差、命令执行、宿主 AI 配置

### 现在可以相信

- **适合人群线索：正在使用 Claude/Codex/Cursor/Gemini 等宿主 AI 的开发者**（supported）：有 supported claim 或项目证据支撑，但仍不等于真实安装效果。 证据：`README.md` Claim：`clm_0002` supported 0.86
- **能力存在：命令行启动或安装流程**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`README.md` Claim：`clm_0001` supported 0.86
- **存在 Quick Start / 安装命令线索**（supported）：可以相信项目文档出现过启动或安装入口；不要因此直接在主力环境运行。 证据：`README.md` Claim：`clm_0003` supported 0.86

### 现在还不能相信

- **角色质量和任务匹配不能直接相信。**（unverified）：角色库证明有很多角色，不证明每个角色都适合你的具体任务，也不证明角色能产生高质量结果。
- **不能把角色文案当成真实执行能力。**（unverified）：安装前只能判断角色描述和任务画像是否匹配，不能证明它能在宿主 AI 里完成任务。
- **真实输出质量不能在安装前相信。**（unverified）：Prompt Preview 只能展示引导方式，不能证明真实项目中的结果质量。
- **宿主 AI 版本兼容性不能在安装前相信。**（unverified）：Claude、Cursor、Codex、Gemini 等宿主加载规则和版本差异必须在真实环境验证。
- **不会污染现有宿主 AI 行为，不能直接相信。**（inferred）：Skill、plugin、AGENTS/CLAUDE/GEMINI 指令可能改变宿主 AI 的默认行为。 证据：`AGENTS.md`
- **可安全回滚不能默认相信。**（unverified）：除非项目明确提供卸载和恢复说明，否则必须先在隔离环境验证。
- **真实安装后是否与用户当前宿主 AI 版本兼容？**（unverified）：兼容性只能通过实际宿主环境验证。
- **项目输出质量是否满足用户具体任务？**（unverified）：安装前预览只能展示流程和边界，不能替代真实评测。

### 继续会触碰什么

- **角色选择偏差**：用户对任务应该由哪个专家角色处理的判断。 原因：选错角色会让 AI 从错误专业视角回答，浪费时间或误导决策。
- **命令执行**：包管理器、网络下载、本地插件目录、项目配置或用户主目录。 原因：运行第一条命令就可能产生环境改动；必须先判断是否值得跑。 证据：`README.md`
- **宿主 AI 配置**：Claude/Codex/Cursor/Gemini/OpenCode 等宿主的 plugin、Skill 或规则加载配置。 原因：宿主配置会改变 AI 后续工作方式，可能和用户已有规则冲突。 证据：`AGENTS.md`
- **本地环境或项目文件**：安装结果、插件缓存、项目配置或本地依赖目录。 原因：安装前无法证明写入范围和回滚方式，需要隔离验证。 证据：`README.md`
- **环境变量 / API Key**：项目入口文档明确出现 API key、token、secret 或账号凭证配置。 原因：如果真实安装需要凭证，应先使用测试凭证并经过权限/合规判断。 证据：`README.md`, `docs/sealos-devbox-guide.md`, `docs/tutorials/getting_started.md`, `examples/example_1_conversation_memory.py` 等
- **宿主 AI 上下文**：AI Context Pack、Prompt Preview、Skill 路由、风险规则和项目事实。 原因：导入上下文会影响宿主 AI 后续判断，必须避免把未验证项包装成事实。

### 最小安全下一步

- **先跑 Prompt Preview**：先用交互式试用验证任务画像和角色匹配，不要先导入整套角色库。（适用：任何项目都适用，尤其是输出质量未知时。）
- **只在隔离目录或测试账号试装**：避免安装命令污染主力宿主 AI、真实项目或用户主目录。（适用：存在命令执行、插件配置或本地写入线索时。）
- **先备份宿主 AI 配置**：Skill、plugin、规则文件可能改变 Claude/Cursor/Codex 的默认行为。（适用：存在插件 manifest、Skill 或宿主规则入口时。）
- **不要使用真实生产凭证**：环境变量/API key 一旦进入宿主或工具链，可能产生账号和合规风险。（适用：出现 API、TOKEN、KEY、SECRET 等环境线索时。）
- **安装后只验证一个最小任务**：先验证加载、兼容、输出质量和回滚，再决定是否深用。（适用：准备从试用进入真实工作流时。）

### 退出方式

- **保留安装前状态**：记录原始宿主配置和项目状态，后续才能判断是否可恢复。
- **准备移除宿主 plugin / Skill / 规则入口**：如果试装后行为异常，可以把宿主 AI 恢复到试装前状态。
- **保留原始角色选择记录**：如果输出偏题，可以回到任务画像阶段重新选择角色，而不是继续沿着错误角色推进。
- **记录安装命令和写入路径**：没有明确卸载说明时，至少要知道哪些目录或配置需要手动清理。
- **准备撤销测试 API key 或 token**：测试凭证泄露或误用时，可以快速止损。
- **如果没有回滚路径，不进入主力环境**：不可回滚是继续前阻断项，不应靠信任或运气继续。

## 哪些只能预览

- 解释项目适合谁和能做什么
- 基于项目文档演示典型对话流程
- 帮助用户判断是否值得安装或继续研究

## 哪些必须安装后验证

- 真实安装 Skill、插件或 CLI
- 执行脚本、修改本地文件或访问外部服务
- 验证真实输出质量、性能和兼容性

## 边界与风险判断卡

- **把安装前预览误认为真实运行**：用户可能高估项目已经完成的配置、权限和兼容性验证。 处理方式：明确区分 prompt_preview_can_do 与 runtime_required。 Claim：`clm_0005` inferred 0.45
- **命令执行会修改本地环境**：安装命令可能写入用户主目录、宿主插件目录或项目配置。 处理方式：先在隔离环境或测试账号中运行。 证据：`README.md` Claim：`clm_0006` supported 0.86
- **待确认**：真实安装后是否与用户当前宿主 AI 版本兼容？。原因：兼容性只能通过实际宿主环境验证。
- **待确认**：项目输出质量是否满足用户具体任务？。原因：安装前预览只能展示流程和边界，不能替代真实评测。
- **待确认**：安装命令是否需要网络、权限或全局写入？。原因：这影响企业环境和个人环境的安装风险。

## 开工前工作上下文

### 加载顺序

- 先读取 how_to_use.host_ai_instruction，建立安装前判断资产的边界。
- 读取 claim_graph_summary，确认事实来自 Claim/Evidence Graph，而不是 Human Wiki 叙事。
- 再读取 intended_users、capabilities 和 quick_start_candidates，判断用户是否匹配。
- 需要执行具体任务时，优先查 role_skill_index，再查 evidence_index。
- 遇到真实安装、文件修改、网络访问、性能或兼容性问题时，转入 risk_card 和 boundaries.runtime_required。

### 任务路由

- **命令行启动或安装流程**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`README.md` Claim：`clm_0001` supported 0.86

### 上下文规模

- 文件总数：223
- 重要文件覆盖：40/223
- 证据索引条目：78
- 角色 / Skill 条目：36

### 证据不足时的处理

- **missing_evidence**：说明证据不足，要求用户提供目标文件、README 段落或安装后验证记录；不要补全事实。
- **out_of_scope_request**：说明该任务超出当前 AI Context Pack 证据范围，并建议用户先查看 Human Manual 或真实安装后验证。
- **runtime_request**：给出安装前检查清单和命令来源，但不要替用户执行命令或声称已执行。
- **source_conflict**：同时展示冲突来源，标记为待核实，不要强行选择一个版本。

## Prompt Recipes

### 适配判断

- 目标：判断这个项目是否适合用户当前任务。
- 预期输出：适配结论、关键理由、证据引用、安装前可预览内容、必须安装后验证内容、下一步建议。

```text
请基于 memu 的 AI Context Pack，先问我 3 个必要问题，然后判断它是否适合我的任务。回答必须包含：适合谁、能做什么、不能做什么、是否值得安装、证据来自哪里。所有项目事实必须引用 evidence_refs、source_paths 或 claim_id。
```

### 安装前体验

- 目标：让用户在安装前感受核心工作流，同时避免把预览包装成真实能力或营销承诺。
- 预期输出：一段带边界标签的体验剧本、安装后验证清单和谨慎建议；不含真实运行承诺或强营销表述。

```text
请把 memu 当作安装前体验资产，而不是已安装工具或真实运行环境。

请严格输出四段：
1. 先问我 3 个必要问题。
2. 给出一段“体验剧本”：用 [安装前可预览]、[必须安装后验证]、[证据不足] 三种标签展示它可能如何引导工作流。
3. 给出安装后验证清单：列出哪些能力只有真实安装、真实宿主加载、真实项目运行后才能确认。
4. 给出谨慎建议：只能说“值得继续研究/试装”“先补充信息后再判断”或“不建议继续”，不得替项目背书。

硬性边界：
- 不要声称已经安装、运行、执行测试、修改文件或产生真实结果。
- 不要写“自动适配”“确保通过”“完美适配”“强烈建议安装”等承诺性表达。
- 如果描述安装后的工作方式，必须使用“如果安装成功且宿主正确加载 Skill，它可能会……”这种条件句。
- 体验剧本只能写成“示例台词/假设流程”：使用“可能会询问/可能会建议/可能会展示”，不要写“已写入、已生成、已通过、正在运行、正在生成”。
- Prompt Preview 不负责给安装命令；如用户准备试装，只能提示先阅读 Quick Start 和 Risk Card，并在隔离环境验证。
- 所有项目事实必须来自 supported claim、evidence_refs 或 source_paths；inferred/unverified 只能作风险或待确认项。

```

### 角色 / Skill 选择

- 目标：从项目里的角色或 Skill 中挑选最匹配的资产。
- 预期输出：候选角色或 Skill 列表，每项包含适用场景、证据路径、风险边界和是否需要安装后验证。

```text
请读取 role_skill_index，根据我的目标任务推荐 3-5 个最相关的角色或 Skill。每个推荐都要说明适用场景、可能输出、风险边界和 evidence_refs。
```

### 风险预检

- 目标：安装或引入前识别环境、权限、规则冲突和质量风险。
- 预期输出：环境、权限、依赖、许可、宿主冲突、质量风险和未知项的检查清单。

```text
请基于 risk_card、boundaries 和 quick_start_candidates，给我一份安装前风险预检清单。不要替我执行命令，只说明我应该检查什么、为什么检查、失败会有什么影响。
```

### 宿主 AI 开工指令

- 目标：把项目上下文转成一次对话开始前的宿主 AI 指令。
- 预期输出：一段边界明确、证据引用明确、适合复制给宿主 AI 的开工前指令。

```text
请基于 memu 的 AI Context Pack，生成一段我可以粘贴给宿主 AI 的开工前指令。这段指令必须遵守 not_runtime=true，不能声称项目已经安装、运行或产生真实结果。
```

## 角色 / Skill 索引

- 共索引 36 个角色 / Skill / 项目文档条目。

- **Architecture Decision Records**（project_doc）：- 0001: Use Workflow Pipelines for Core Operations 0001-workflow-pipeline-architecture.md - 0002: Use Pluggable Storage with Backend-Specific Vector Search 0002-pluggable-storage-and-vector-strategy.md - 0003: Model User Scope as First-Class Fields on Memory Records 0003-user-scope-in-data-model.md - 0004: Add Workspace Memorize Without Touching Single-File memorize 0004-workspace-memorize-and-memory-file-system.md 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/adr/README.md`
- **AGENTS.md**（project_doc）：Guidance for AI coding agents working in this repository. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`AGENTS.md`
- **memU**（project_doc）：Every agent needs a workspace runtime 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`README.md`
- **MemU Assistant - Sealos DevBox Example**（project_doc）：MemU Assistant - Sealos DevBox Example 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/sealos-assistant/README.md`
- **Contributing to MemU**（project_doc）：Thank you for your interest in contributing to MemU! This document provides guidelines and information for contributors. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`CONTRIBUTING.md`
- **GitHub Issue Draft: Memory Types + Tool Memory**（project_doc）：GitHub Issue Draft: Memory Types + Tool Memory 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/HACKATHON_ISSUE_DRAFT.md`
- **🔥🔥🔥 MAD COMBO OPTIONS FOR MEMU HACKATHON 🔥🔥🔥**（project_doc）：🔥🔥🔥 MAD COMBO OPTIONS FOR MEMU HACKATHON 🔥🔥🔥 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/HACKATHON_MAD_COMBOS.md`
- **memU Architecture**（project_doc）：This document describes the self-hosted memu Python package architecture as implemented in this repository. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/architecture.md`
- **MemU LangGraph Integration**（project_doc）：The MemU LangGraph Integration provides a seamless adapter to expose MemU's powerful memory capabilities memorize and retrieve as standard LangChain https://python.langchain.com/ / LangGraph https://langchain-ai.github.io/langgraph/ tools. This allows your agents to persist information and recall it across sessions using MemU as the long-term memory backend. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/langgraph_integration.md`
- **Deploying MemU on Sealos DevBox**（project_doc）：This guide demonstrates how to build and deploy a Personal AI Assistant with Long-Term Memory using MemU on Sealos DevBox https://sealos.io/products/devbox . 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/sealos-devbox-guide.md`
- **🛡️ Context-Aware Support Agent Sealos Edition**（project_doc）：🛡️ Context-Aware Support Agent Sealos Edition 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/sealos_use_case.md`
- **SQLite Database Integration**（project_doc）：MemU supports SQLite as a lightweight, file-based database backend for memory storage. This is ideal for: 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/sqlite.md`
- **ADR 0001: Use Workflow Pipelines for Core Operations**（project_doc）：ADR 0001: Use Workflow Pipelines for Core Operations 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/adr/0001-workflow-pipeline-architecture.md`
- **ADR 0002: Use Pluggable Storage with Backend-Specific Vector Search**（project_doc）：ADR 0002: Use Pluggable Storage with Backend-Specific Vector Search 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/adr/0002-pluggable-storage-and-vector-strategy.md`
- **ADR 0003: Model User Scope as First-Class Fields on Memory Records**（project_doc）：ADR 0003: Model User Scope as First-Class Fields on Memory Records 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/adr/0003-user-scope-in-data-model.md`
- **ADR 0004: Add Workspace Memorize Without Touching Single-File memorize**（project_doc）：ADR 0004: Add Workspace Memorize Without Touching Single-File memorize 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/adr/0004-workspace-memorize-and-memory-file-system.md`
- **Grok xAI Integration**（project_doc）：MemU supports Grok , the AI model from xAI, as a first-class LLM provider. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/integrations/grok.md`
- **Grok xAI Provider**（project_doc）：memU includes first-class support for Grok https://grok.x.ai/ , allowing you to leverage xAI's powerful language models directly within your application. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/providers/grok.md`
- **Quickstart: Adding Long-Term Memory to Python Agents**（project_doc）：Quickstart: Adding Long-Term Memory to Python Agents 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/tutorials/getting_started.md`
- **Changelog**（project_doc）：1.5.1 https://github.com/NevaMind-AI/memU/compare/v1.5.0...v1.5.1 2026-03-23 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`CHANGELOG.md`
- **memU**（project_doc）：File System as Memory, Memory Shapes the Agent 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_en.md`
- **memU**（project_doc）：El sistema de archivos como memoria, la memoria moldea al agente 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_es.md`
- **memU**（project_doc）：Le système de fichiers comme mémoire, la mémoire façonne l'agent 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_fr.md`
- **memU**（project_doc）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_ja.md`
- **memU**（project_doc）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_ko.md`
- **memU**（project_doc）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`readme/README_zh.md`
- **activities**（project_doc）：activities Open Source Contributions - The user enjoys contributing to open source projects, specifically a Python CLI tool used for automating deployment tasks. Running - The user usually goes for a run every morning and is interested in running routes near downtown San Francisco. Dining Plans - The user plans to try several vegetarian restaurants in San Francisco for their partner, who is vegetarian. Gym - The use… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/activities.md`
- **experiences**（project_doc）：experiences User Experiences - The user has been programming for about 5 years. - The user is leading a big product launch next month, which is causing work-related stress that is affecting their sleep schedule. - The user usually goes to bed around 11 PM but finds themselves awake thinking about work projects. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/experiences.md`
- **goals**（project_doc）：goals Learning Objectives - The user is interested in learning more about system design and scalability patterns. - The user is learning about event-driven architecture and message queues, specifically using Apache Kafka for event streaming. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/goals.md`
- **habits**（project_doc）：habits Eating Habits - The user is trying to eat less meat. Exercise Routine - The user usually goes for a run every morning. - The user exercises regularly, going to the gym 3-4 times a week. Sleep Habits - The user has been having trouble sleeping due to work stress. - The user usually tries to go to bed around 11 PM but struggles to fall asleep due to stress from work projects. - The user checks their phone befor… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/habits.md`
- **knowledge**（project_doc）：knowledge Reading Interests - The user has been reading about OpenAPI specifications and exploring tools like Swagger and Postman. Technology Stack - Alex's technology stack includes Django and FastAPI for Python services, Kubernetes for orchestration, Redis for caching, and Apache Kafka for event streaming. Interests - Alex has an interest in system design and scalability patterns and is learning about event-driven… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/knowledge.md`
- **Opinions**（project_doc）： 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/opinions.md`
- **personal info**（project_doc）：personal info Basic Information - The user is named Alex and works as a software engineer at TechCorp. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/personal_info.md`
- **preferences**（project_doc）：preferences Interests - The user loves food and nature. - The user enjoys exploring tech companies due to their background in software development. - The user is interested in vegetarian options as their partner is vegetarian and the user is trying to eat less meat. - The user likes reading and used to read before bed. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/preferences.md`
- **relationships**（project_doc）：relationships partner - The user's partner enjoys photography and museums. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/relationships.md`
- **work life**（project_doc）：work life Professional Background - The user is a software engineer at TechCorp. - Alex is a software engineer with approximately 5 years of programming experience. - The user works in software development. Current Responsibilities - The user primarily works on backend systems using Python and Go for an e-commerce platform. - At TechCorp, Alex's team is building a distributed microservices architecture for their e-c… 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`examples/output/conversation_example/work_life.md`

## 证据索引

- 共索引 78 条证据。

- **Architecture Decision Records**（documentation）：- 0001: Use Workflow Pipelines for Core Operations 0001-workflow-pipeline-architecture.md - 0002: Use Pluggable Storage with Backend-Specific Vector Search 0002-pluggable-storage-and-vector-strategy.md - 0003: Model User Scope as First-Class Fields on Memory Records 0003-user-scope-in-data-model.md - 0004: Add Workspace Memorize Without Touching Single-File memorize 0004-workspace-memorize-and-memory-file-system.md 证据：`docs/adr/README.md`
- **AGENTS.md**（documentation）：Guidance for AI coding agents working in this repository. 证据：`AGENTS.md`
- **memU**（documentation）：Every agent needs a workspace runtime 证据：`README.md`
- **MemU Assistant - Sealos DevBox Example**（documentation）：MemU Assistant - Sealos DevBox Example 证据：`examples/sealos-assistant/README.md`
- **Contributing to MemU**（documentation）：Thank you for your interest in contributing to MemU! This document provides guidelines and information for contributors. 证据：`CONTRIBUTING.md`
- **GitHub Issue Draft: Memory Types + Tool Memory**（documentation）：GitHub Issue Draft: Memory Types + Tool Memory 证据：`docs/HACKATHON_ISSUE_DRAFT.md`
- **🔥🔥🔥 MAD COMBO OPTIONS FOR MEMU HACKATHON 🔥🔥🔥**（documentation）：🔥🔥🔥 MAD COMBO OPTIONS FOR MEMU HACKATHON 🔥🔥🔥 证据：`docs/HACKATHON_MAD_COMBOS.md`
- **memU Architecture**（documentation）：This document describes the self-hosted memu Python package architecture as implemented in this repository. 证据：`docs/architecture.md`
- **MemU LangGraph Integration**（documentation）：The MemU LangGraph Integration provides a seamless adapter to expose MemU's powerful memory capabilities memorize and retrieve as standard LangChain https://python.langchain.com/ / LangGraph https://langchain-ai.github.io/langgraph/ tools. This allows your agents to persist information and recall it across sessions using MemU as the long-term memory backend. 证据：`docs/langgraph_integration.md`
- **Deploying MemU on Sealos DevBox**（documentation）：This guide demonstrates how to build and deploy a Personal AI Assistant with Long-Term Memory using MemU on Sealos DevBox https://sealos.io/products/devbox . 证据：`docs/sealos-devbox-guide.md`
- **🛡️ Context-Aware Support Agent Sealos Edition**（documentation）：🛡️ Context-Aware Support Agent Sealos Edition 证据：`docs/sealos_use_case.md`
- **SQLite Database Integration**（documentation）：MemU supports SQLite as a lightweight, file-based database backend for memory storage. This is ideal for: 证据：`docs/sqlite.md`
- **ADR 0001: Use Workflow Pipelines for Core Operations**（documentation）：ADR 0001: Use Workflow Pipelines for Core Operations 证据：`docs/adr/0001-workflow-pipeline-architecture.md`
- **ADR 0002: Use Pluggable Storage with Backend-Specific Vector Search**（documentation）：ADR 0002: Use Pluggable Storage with Backend-Specific Vector Search 证据：`docs/adr/0002-pluggable-storage-and-vector-strategy.md`
- **ADR 0003: Model User Scope as First-Class Fields on Memory Records**（documentation）：ADR 0003: Model User Scope as First-Class Fields on Memory Records 证据：`docs/adr/0003-user-scope-in-data-model.md`
- **ADR 0004: Add Workspace Memorize Without Touching Single-File memorize**（documentation）：ADR 0004: Add Workspace Memorize Without Touching Single-File memorize 证据：`docs/adr/0004-workspace-memorize-and-memory-file-system.md`
- **Grok xAI Integration**（documentation）：MemU supports Grok , the AI model from xAI, as a first-class LLM provider. 证据：`docs/integrations/grok.md`
- **Grok xAI Provider**（documentation）：memU includes first-class support for Grok https://grok.x.ai/ , allowing you to leverage xAI's powerful language models directly within your application. 证据：`docs/providers/grok.md`
- **Quickstart: Adding Long-Term Memory to Python Agents**（documentation）：Quickstart: Adding Long-Term Memory to Python Agents 证据：`docs/tutorials/getting_started.md`
- **Changelog**（documentation）：1.5.1 https://github.com/NevaMind-AI/memU/compare/v1.5.0...v1.5.1 2026-03-23 证据：`CHANGELOG.md`
- **memU**（documentation）：File System as Memory, Memory Shapes the Agent 证据：`readme/README_en.md`
- **memU**（documentation）：El sistema de archivos como memoria, la memoria moldea al agente 证据：`readme/README_es.md`
- **memU**（documentation）：Le système de fichiers comme mémoire, la mémoire façonne l'agent 证据：`readme/README_fr.md`
- **memU**（documentation）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord.com/invite/hQZntfGsbJ ! Twitter https://img.shields.io/badge/Twitter-Follow-1DA1F2?logo=x&logoColor=white https://x.com/memU ai 证据：`readme/README_ja.md`
- **memU**（documentation）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord.com/invite/hQZntfGsbJ ! Twitter https://img.shields.io/badge/Twitter-Follow-1DA1F2?logo=x&logoColor=white https://x.com/memU ai 证据：`readme/README_ko.md`
- **memU**（documentation）：! PyPI version https://badge.fury.io/py/memu-py.svg https://badge.fury.io/py/memu-py ! License: Apache 2.0 https://img.shields.io/badge/License-Apache%202.0-blue.svg https://opensource.org/licenses/Apache-2.0 ! Python 3.13+ https://img.shields.io/badge/python-3.13+-blue.svg https://www.python.org/downloads/ ! Discord https://img.shields.io/badge/Discord-Join%20Chat-5865F2?logo=discord&logoColor=white https://discord.com/invite/hQZntfGsbJ ! Twitter https://img.shields.io/badge/Twitter-Follow-1DA1F2?logo=x&logoColor=white https://x.com/memU ai 证据：`readme/README_zh.md`
- **activities**（documentation）：activities Open Source Contributions - The user enjoys contributing to open source projects, specifically a Python CLI tool used for automating deployment tasks. Running - The user usually goes for a run every morning and is interested in running routes near downtown San Francisco. Dining Plans - The user plans to try several vegetarian restaurants in San Francisco for their partner, who is vegetarian. Gym - The user goes to the gym 3-4 times a week, usually after work around 7 PM. 证据：`examples/output/conversation_example/activities.md`
- **experiences**（documentation）：experiences User Experiences - The user has been programming for about 5 years. - The user is leading a big product launch next month, which is causing work-related stress that is affecting their sleep schedule. - The user usually goes to bed around 11 PM but finds themselves awake thinking about work projects. 证据：`examples/output/conversation_example/experiences.md`
- **goals**（documentation）：goals Learning Objectives - The user is interested in learning more about system design and scalability patterns. - The user is learning about event-driven architecture and message queues, specifically using Apache Kafka for event streaming. 证据：`examples/output/conversation_example/goals.md`
- **habits**（documentation）：habits Eating Habits - The user is trying to eat less meat. Exercise Routine - The user usually goes for a run every morning. - The user exercises regularly, going to the gym 3-4 times a week. Sleep Habits - The user has been having trouble sleeping due to work stress. - The user usually tries to go to bed around 11 PM but struggles to fall asleep due to stress from work projects. - The user checks their phone before bed, which may impact their sleep quality. - The user used to read before bed but stopped because they were always checking work emails. Caffeine Consumption - The user usually drinks coffee throughout the day to stay alert, typically having their last coffee around 3-4 PM. 证据：`examples/output/conversation_example/habits.md`
- **knowledge**（documentation）：knowledge Reading Interests - The user has been reading about OpenAPI specifications and exploring tools like Swagger and Postman. Technology Stack - Alex's technology stack includes Django and FastAPI for Python services, Kubernetes for orchestration, Redis for caching, and Apache Kafka for event streaming. Interests - Alex has an interest in system design and scalability patterns and is learning about event-driven architecture and message queues. - For food in San Francisco, check out the Ferry Building Marketplace for gourmet options and local produce. - Golden Gate Park, Lands End, and Muir Woods are recommended for nature activities in San Francisco. - A self-guided tour of tech headqu… 证据：`examples/output/conversation_example/knowledge.md`
- **Opinions**（documentation）：No content available 证据：`examples/output/conversation_example/opinions.md`
- **personal info**（documentation）：personal info Basic Information - The user is named Alex and works as a software engineer at TechCorp. 证据：`examples/output/conversation_example/personal_info.md`
- **preferences**（documentation）：preferences Interests - The user loves food and nature. - The user enjoys exploring tech companies due to their background in software development. - The user is interested in vegetarian options as their partner is vegetarian and the user is trying to eat less meat. - The user likes reading and used to read before bed. 证据：`examples/output/conversation_example/preferences.md`
- **relationships**（documentation）：relationships partner - The user's partner enjoys photography and museums. 证据：`examples/output/conversation_example/relationships.md`
- **work life**（documentation）：work life Professional Background - The user is a software engineer at TechCorp. - Alex is a software engineer with approximately 5 years of programming experience. - The user works in software development. Current Responsibilities - The user primarily works on backend systems using Python and Go for an e-commerce platform. - At TechCorp, Alex's team is building a distributed microservices architecture for their e-commerce platform. - The user uses Django and FastAPI for Python services and is migrating to Go for better performance. - The user works with Kubernetes for orchestration and Redis for caching. - The user is responsible for monitoring and observability, using Prometheus and Grafa… 证据：`examples/output/conversation_example/work_life.md`
- **Conv1**（structured_config）：{ "role": "user", "content": "Hi! I'm Alex. I'm a software engineer working at TechCorp. I've been programming for about 5 years now." }, { "role": "assistant", "content": "Nice to meet you, Alex! It's great to have you here. What kind of software development do you primarily work on at TechCorp?" }, { "role": "user", "content": "I mainly work on backend systems using Python and Go. We're building a distributed microservices architecture for our e-commerce platform. I'm particularly interested in learning more about system design and scalability patterns." }, { "role": "assistant", "content": "That sounds like an exciting project! Distributed systems and microservices certainly present inte… 证据：`examples/resources/conversations/conv1.json`
- **Conv2**（structured_config）：{ "role": "user", "created at": "2025-02-01T09:00:00Z", "content": "I need help planning a weekend trip to San Francisco. Can you suggest some activities?" }, { "role": "assistant", "created at": "2025-02-01T09:00:08Z", "content": "I'd be happy to help you plan your San Francisco trip! What are your main interests? Are you into tech, food, nature, art, or something else?" }, { "role": "user", "created at": "2025-02-01T09:00:25Z", "content": "I love food and nature! I'm also interested in visiting some tech companies since I work in software development. My partner enjoys photography and museums." }, { "role": "assistant", "created at": "2025-02-01T09:00:40Z", "content": "Perfect! For food,… 证据：`examples/resources/conversations/conv2.json`
- **Conv3**（structured_config）：{ "role": "user", "created at": "2025-03-10T22:00:00Z", "content": "I've been having trouble sleeping lately. I think work stress might be affecting my sleep schedule." }, { "role": "assistant", "created at": "2025-03-10T22:00:12Z", "content": "I'm sorry to hear you're having trouble sleeping. Work stress can definitely impact sleep quality. Can you tell me more about your current sleep routine and what's been going on at work?" }, { "role": "user", "created at": "2025-03-10T22:00:45Z", "content": "I usually try to go to bed around 11 PM, but I find myself lying awake thinking about work projects. I'm leading a big product launch next month and there's a lot of pressure. I also tend to chec… 证据：`examples/resources/conversations/conv3.json`
- **Byte-compiled / optimized / DLL files**（source_file）：data/ Byte-compiled / optimized / DLL files pycache / .py cod $py.class 证据：`.gitignore`
- **.Pre Commit Config**（source_file）：repos: - repo: https://github.com/pre-commit/pre-commit-hooks rev: "v6.0.0" hooks: - id: check-case-conflict - id: check-merge-conflict - id: check-toml - id: check-yaml - id: check-json - id: pretty-format-json args: --autofix, --no-sort-keys - id: end-of-file-fixer - id: trailing-whitespace - repo: https://github.com/astral-sh/ruff-pre-commit rev: "v0.14.3" hooks: - id: ruff args: --exit-non-zero-on-fix - id: ruff-format 证据：`.pre-commit-config.yaml`
- **.python-version**（source_file）：3.13 证据：`.python-version`
- **"cdylib" is necessary to produce a shared library for Python to import from.**（source_file）：package name = "memu" version = "0.1.0" edition = "2024" 证据：`Cargo.toml`
- **License**（source_file）：Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ 证据：`LICENSE.txt`
- **Manifest**（source_file）：include README.md recursive-include memu .py prune example include setup postgres env.sh exclude .env exclude .env.example exclude setup.py.backup exclude / pycache / exclude .pyc exclude .pyo exclude .git/ exclude .github/ exclude server/ exclude docs/ exclude scripts/ exclude docker-compose.yml exclude Dockerfile exclude .dockerignore exclude PROJECT RELEASE SUMMARY.md exclude Makefile exclude .pre-commit-config.yaml 证据：`MANIFEST.in`
- **Makefile**（source_file）：.PHONY: install install: @echo "🚀 Creating virtual environment using uv" @uv sync @uv run pre-commit install 证据：`Makefile`
- **Write to output files**（source_file）：src path = os.path.abspath "src" ⋮---- async def generate memory md categories, output dir ⋮---- generated files = ⋮---- name = cat.get "name", "unknown" summary = cat.get "summary", "" filename = f"{name}.md" filepath = os.path.join output dir, filename ⋮---- cleaned summary = summary.replace " ", "" .replace " ", "" .strip ⋮---- async def main ⋮---- api key = os.getenv "OPENAI API KEY" ⋮---- msg = "Please set OPENAI API KEY environment variable" ⋮---- service = MemoryService conversation files = ⋮---- total items = 0 categories = ⋮---- result = await service.memorize resource url=conv file, modality="conversation" ⋮---- categories = result.get "categories", ⋮---- Write to output files out… 证据：`examples/example_1_conversation_memory.py`
- **Construct prompt for LLM**（source_file）：src path = os.path.abspath "src" ⋮---- categories with content = cat for cat in categories if cat.get "summary" and cat.get "summary" .strip ⋮---- categories text = "\n\n".join Construct prompt for LLM prompt = f"""Generate a concise production-ready task execution guide. client = AsyncOpenAI api key=service.llm config.api key response = await client.chat.completions.create generated content = response.choices 0 .message.content Write to file ⋮---- async def main 证据：`examples/example_2_skill_extraction.py`
- **Write to output files**（source_file）：src path = os.path.abspath "src" ⋮---- async def generate memory md categories, output dir ⋮---- generated files = ⋮---- name = cat.get "name", "unknown" description = cat.get "description", "" summary = cat.get "summary", "" filename = f"{name}.md" filepath = os.path.join output dir, filename ⋮---- formatted name = name.replace " ", " " .title ⋮---- cleaned summary = summary.replace " ", "" .replace " ", "" .strip ⋮---- async def main ⋮---- api key = os.getenv "OPENAI API KEY" ⋮---- msg = "Please set OPENAI API KEY environment variable" ⋮---- multimodal categories = service = MemoryService resources = ⋮---- total items = 0 categories = ⋮---- result = await service.memorize resource url=res… 证据：`examples/example_3_multimodal_memory.py`
- **Example 4 Openrouter Memory**（source_file）：src path = os.path.abspath "src" ⋮---- async def generate memory md categories, output dir ⋮---- generated files = ⋮---- name = cat.get "name", "unknown" summary = cat.get "summary", "" filename = f"{name}.md" filepath = os.path.join output dir, filename ⋮---- cleaned summary = summary.replace " ", "" .replace " ", "" .strip ⋮---- async def main ⋮---- api key = os.getenv "OPENROUTER API KEY" ⋮---- msg = "Please set OPENROUTER API KEY environment variable" ⋮---- service = MemoryService conversation files = ⋮---- total items = 0 categories = ⋮---- result = await service.memorize resource url=conv file, modality="conversation" ⋮---- categories = result.get "categories", ⋮---- output dir = "exa… 证据：`examples/example_4_openrouter_memory.py`
- **Output generation**（source_file）：project root = Path file .parent.parent src path = str project root / "src" ⋮---- async def run conversation memory demo service ⋮---- conversation files = total items = 0 categories = ⋮---- result = await service.memorize resource url=conv file, modality="conversation" ⋮---- categories = result.get "categories", ⋮---- Output generation output dir = "examples/output/lazyllm example/conversation" ⋮---- ========================================== PART 2: Skill Extraction ⋮---- async def run skill extraction demo service ⋮---- skill prompt = """ ⋮---- logs = "examples/resources/logs/log1.txt", "examples/resources/logs/log2.txt", "examples/resources/logs/log3.txt" all skills = ⋮---- result = awa… 证据：`examples/example_5_with_lazyllm_client.py`
- **Getting Started Robust**（source_file）：async def main - None ⋮---- api key = os.getenv "OPENAI API KEY" ⋮---- service = MemoryService ⋮---- memory content = "The user is a senior Python architect who loves clean code and type hints." result = await service.create memory item ⋮---- query text = "What kind of code does the user like?" ⋮---- search results = await service.retrieve queries= {"role": "user", "content": query text} items = search results.get "items", 证据：`examples/getting_started_robust.py`
- **Langgraph Demo**（source_file）：logger = logging.getLogger "langgraph demo" async def initialize infrastructure - MemULangGraphTools ⋮---- service = MemoryService ⋮---- async def process conversation tools: list BaseTool , user id: str - None ⋮---- save tool = next t for t in tools if t.name == "save memory" ⋮---- inputs = { result = await save tool.ainvoke inputs ⋮---- async def process retrieval tools: list BaseTool , user id: str - None ⋮---- search tool = next t for t in tools if t.name == "search memory" ⋮---- inputs = {"query": "What are the user's preferences?", "user id": user id, "limit": 3} result = await search tool.ainvoke inputs ⋮---- async def main - None ⋮---- adapter = await initialize infrastructure tools… 证据：`examples/langgraph_demo.py`
- **2. MEMORY INGESTION PHASE 1**（source_file）：MEMU INSTALLED = True ⋮---- MEMU INSTALLED = False IMPORT ERROR = str e def print slow text, delay=0.02 def run rigorous demo ⋮---- 2. MEMORY INGESTION PHASE 1 ⋮---- response = '🤖 Agent: "Welcome back, Captain. Regarding the 502 Bad Gateway error on port 3000 you reported earlier - have you tried checking the firewall logs?"' 证据：`examples/sealos_support_agent.py`
- **license = {file = "LICENSE"}**（source_file）：project name = "memu-py" version = "1.5.1" authors = {name = "MemU Team", email = "contact@nevamind.ai"}, description = "AI Memory and Conversation Management Framework - Simple as mem0, Powerful as MemU" readme = "README.md" license = {file = "LICENSE"} requires-python = " =3.13" classifiers = "Development Status :: 4 - Beta", "Intended Audience :: Developers", 证据：`pyproject.toml`
- **Setup**（source_file）：flake8 max-line-length = 120 extend-ignore = E203,W503,E501 exclude = .git, pycache , .venv, venv, build, dist, .egg-info, .pytest cache, .mypy cache per-file-ignores = /test .py:E402 /test .py:E402 /tests.py:E402 /quick memory test.py:E402 证据：`setup.cfg`
- **Lib**（source_file）：fn hello from bin - String { "Hello from memu!".to string ⋮---- fn core m: &Bound - PyResult { m.add function wrap pyfunction! hello from bin, m ? ?; Ok 证据：`src/lib.rs`
- **Config**（source_file）：memorize config = { retrieve config = { 证据：`examples/proactive/memory/config.py`
- **Common**（source_file）：USER ID = "claude user" SHARED MEMORY SERVICE = None def get memory service - MemoryService ⋮---- api key = os.getenv "OPENAI API KEY" ⋮---- msg = "Please set OPENAI API KEY environment variable" ⋮---- SHARED MEMORY SERVICE = MemoryService 证据：`examples/proactive/memory/local/common.py`
- **Memorize**（source_file）：USER ID = "claude user" ⋮---- resource data = { time string = pendulum.now .format "YYYYMMDD HHmmss" resource url = Path file .parent / "data" / f"conv {time string}.json" ⋮---- def memorize conversation messages: list dict str, Any - Awaitable dict str, Any ⋮---- memory service = get memory service resource url = dump conversation resource conversation messages 证据：`examples/proactive/memory/local/memorize.py`
- 其余 18 条证据见 `AI_CONTEXT_PACK.json` 或 `EVIDENCE_INDEX.json`。

## 宿主 AI 必须遵守的规则

- **把本资产当作开工前上下文，而不是运行环境。**：AI Context Pack 只包含证据化项目理解，不包含目标项目的可执行状态。 证据：`docs/adr/README.md`, `AGENTS.md`, `README.md`
- **回答用户时区分可预览内容与必须安装后才能验证的内容。**：安装前体验的消费者价值来自降低误装和误判，而不是伪装成真实运行。 证据：`docs/adr/README.md`, `AGENTS.md`, `README.md`

## 用户开工前应该回答的问题

- 你准备在哪个宿主 AI 或本地环境中使用它？
- 你只是想先体验工作流，还是准备真实安装？
- 你最在意的是安装成本、输出质量、还是和现有规则的冲突？

## 验收标准

- 所有能力声明都能回指到 evidence_refs 中的文件路径。
- AI_CONTEXT_PACK.md 没有把预览包装成真实运行。
- 用户能在 3 分钟内看懂适合谁、能做什么、如何开始和风险边界。

---

## Doramagic Context Augmentation

下面内容用于强化 Repomix/AI Context Pack 主体。Human Manual 只提供阅读骨架；踩坑日志会被转成宿主 AI 必须遵守的工作约束。

## Human Manual 骨架

使用规则：这里只是项目阅读路线和显著性信号，不是事实权威。具体事实仍必须回到 repo evidence / Claim Graph。

宿主 AI 硬性规则：
- 不得把页标题、章节顺序、摘要或 importance 当作项目事实证据。
- 解释 Human Manual 骨架时，必须明确说它只是阅读路线/显著性信号。
- 能力、安装、兼容性、运行状态和风险判断必须引用 repo evidence、source path 或 Claim Graph。

- **memU 概览与系统架构**：importance `high`
  - source_paths: README.md, docs/architecture.md, src/memu/workflow/pipeline.py, src/memu/workflow/runner.py, src/memu/workflow/step.py
- **核心数据模型与 memorize/retrieve API**：importance `high`
  - source_paths: src/memu/app/memorize.py, src/memu/app/retrieve.py, src/memu/app/crud.py, src/memu/database/models.py, src/memu/database/repositories/memory_item.py
- **可插拔存储后端与 LLM/Embedding 路由**：importance `high`
  - source_paths: src/memu/database/factory.py, src/memu/database/inmemory/repo.py, src/memu/database/inmemory/vector.py, src/memu/database/sqlite/sqlite.py, src/memu/database/sqlite/models.py
- **部署、扩展、常见故障与已知问题**：importance `high`
  - source_paths: pyproject.toml, Makefile, src/memu/integrations/langgraph.py, src/memu/blob/local_fs.py, src/memu/blob/folder.py

## Repo Inspection Evidence / 源码检查证据

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `6824b541be4274bf856b2f80e960163273d62227`
- inspected_files: `README.md`, `pyproject.toml`, `uv.lock`, `docs/HACKATHON_ISSUE_DRAFT.md`, `docs/HACKATHON_MAD_COMBOS.md`, `docs/adr/0001-workflow-pipeline-architecture.md`, `docs/adr/0002-pluggable-storage-and-vector-strategy.md`, `docs/adr/0003-user-scope-in-data-model.md`, `docs/adr/0004-workspace-memorize-and-memory-file-system.md`, `docs/adr/README.md`, `docs/architecture.md`, `docs/integrations/grok.md`, `docs/langgraph_integration.md`, `docs/providers/grok.md`, `docs/sealos-devbox-guide.md`, `docs/sealos_use_case.md`, `docs/sqlite.md`, `docs/tutorials/getting_started.md`, `examples/example_1_conversation_memory.py`, `examples/example_2_skill_extraction.py`

宿主 AI 硬性规则：
- 没有 repo_clone_verified=true 时，不得声称已经读过源码。
- 没有 repo_inspection_verified=true 时，不得把 README/docs/package 文件判断写成事实。
- 没有 quick_start_verified=true 时，不得声称 Quick Start 已跑通。

## Doramagic Pitfall Constraints / 踩坑约束

这些规则来自 Doramagic 发现、验证或编译过程中的项目专属坑点。宿主 AI 必须把它们当作工作约束，而不是普通说明文字。

### Constraint 1: 来源证据：[BUG] memory_items table's happend_at collumn stores NULL value even though when proper timestamp provided in messages.

- Trigger: GitHub 社区证据显示该项目存在一个配置相关的待验证问题：[BUG] memory_items table's happend_at collumn stores NULL value even though when proper timestamp provided in messages.
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | https://github.com/NevaMind-AI/memU/issues/428 | 来源类型 github_issue 暴露的待验证使用条件。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 2: 来源证据：Bug: memu-server entry point points to missing module (memu.server.cli)

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：Bug: memu-server entry point points to missing module (memu.server.cli)
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | https://github.com/NevaMind-AI/memU/issues/354 | 来源讨论提到 python 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 3: 来源证据：[BUG] sqlite backend embding issue

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：[BUG] sqlite backend embding issue
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | https://github.com/NevaMind-AI/memU/issues/382 | 来源类型 github_issue 暴露的待验证使用条件。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 4: 可能修改宿主 AI 配置

- Trigger: 项目面向 Claude/Cursor/Codex/Gemini/OpenCode 等宿主，或安装命令涉及用户配置目录。
- Host AI rule: 列出会写入的配置文件、目录和卸载/回滚步骤。
- Why it matters: 安装可能改变本机 AI 工具行为，用户需要知道写入位置和回滚方法。
- Evidence: capability.host_targets | https://github.com/NevaMind-AI/memU | host_targets=openclaw, mcp_host, claude
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 5: 能力判断依赖假设

- Trigger: README/documentation is current enough for a first validation pass.
- Host AI rule: 将假设转成下游验证清单。
- Why it matters: 假设不成立时，用户拿不到承诺的能力。
- Evidence: capability.assumptions | https://github.com/NevaMind-AI/memU | README/documentation is current enough for a first validation pass.
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 6: 维护活跃度未知

- Trigger: 未记录 last_activity_observed。
- Host AI rule: 补 GitHub 最近 commit、release、issue/PR 响应信号。
- Why it matters: 新项目、停更项目和活跃项目会被混在一起，推荐信任度下降。
- Evidence: evidence.maintainer_signals | https://github.com/NevaMind-AI/memU | last_activity_observed missing
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

- Trigger: no_demo
- Evidence: downstream_validation.risk_items | https://github.com/NevaMind-AI/memU | no_demo; severity=medium
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 8: 存在评分风险

- Trigger: no_demo
- Why it matters: 风险会影响是否适合普通用户安装。
- Evidence: risks.scoring_risks | https://github.com/NevaMind-AI/memU | no_demo; severity=medium
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 9: issue/PR 响应质量未知

- Trigger: issue_or_pr_quality=unknown。
- Host AI rule: 抽样最近 issue/PR，判断是否长期无人处理。
- Why it matters: 用户无法判断遇到问题后是否有人维护。
- Evidence: evidence.maintainer_signals | https://github.com/NevaMind-AI/memU | issue_or_pr_quality=unknown
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 10: 发布节奏不明确

- Trigger: release_recency=unknown。
- Host AI rule: 确认最近 release/tag 和 README 安装命令是否一致。
- Why it matters: 安装命令和文档可能落后于代码，用户踩坑概率升高。
- Evidence: evidence.maintainer_signals | https://github.com/NevaMind-AI/memU | release_recency=unknown
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。