@pcircle/memesh 4.1.3 → 4.1.4

package/.mcp.json

@@ -1,9 +1,7 @@
 {
   "mcpServers": {
     "memesh": {
-      "command": "node",
-      "args": ["${CLAUDE_PLUGIN_ROOT}/dist/mcp/server.js"],
-      "env": { "NODE_ENV": "production" }
+      "command": "memesh-mcp"
     }
   }
 }

package/README.de.md

@@ -37,6 +37,17 @@ Dieses Paket ist die lokale Speicherschicht der MeMesh-Produktfamilie. Es ist be
 npm install -g @pcircle/memesh
 ```
 
+### Schritt 1.5: MeMesh in Claude Code einbinden (empfohlen, einmalig)
+
+`npm install -g` legt die CLI in den PATH und registriert den MCP-Server, aber MeMeshs Claude-Code-Session-Hooks werden **nicht** automatisch verdrahtet. Ohne diese Hooks können Sie `memesh remember` / `recall` manuell verwenden, aber die **Auto-Capture-Schleife** (Session → Lektionen → proaktive Erinnerung in der nächsten Session) bleibt stumm.
+
+```bash
+memesh install-hooks         # fügt memesh-Hooks zu ~/.claude/settings.json hinzu
+memesh doctor                # bestätigt, dass „Hooks wired into Claude Code" PASST
+```
+
+Die Hooks existieren neben Ihren bestehenden Custom-Hooks unter `~/.claude/hooks/` — `install-hooks` schreibt additiv und überschreibt nie Ihre Einträge. Zum Entfernen: `memesh uninstall-hooks`.
+
 ### Schritt 2: Entscheidung speichern
 
 ```bash
@@ -271,7 +282,7 @@ Der Kern ist Framework-agnostisch. Dieselbe Logik läuft vom Terminal, HTTP oder
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.es.md

@@ -37,6 +37,17 @@ Este paquete es la capa de memoria local de la familia de productos MeMesh. Es i
 npm install -g @pcircle/memesh
 ```
 
+### Paso 1.5: Conecta MeMesh a Claude Code (recomendado, una sola vez)
+
+`npm install -g` pone la CLI en el PATH y registra el servidor MCP, pero **no** conecta automáticamente los hooks de sesión de MeMesh a Claude Code. Sin estos hooks, puedes usar `memesh remember` / `recall` manualmente, pero el **bucle de captura automática** (sesión → lecciones → recall proactivo en la siguiente sesión) queda en silencio.
+
+```bash
+memesh install-hooks         # añade los hooks de memesh a ~/.claude/settings.json
+memesh doctor                # confirma que "Hooks wired into Claude Code" pasa
+```
+
+Estos hooks coexisten con tus hooks personalizados en `~/.claude/hooks/` — `install-hooks` escribe de forma aditiva y nunca sobrescribe los tuyos. Para eliminarlos: `memesh uninstall-hooks`.
+
 ### Paso 2: Guarda una decisión
 
 ```bash
@@ -146,13 +157,14 @@ Pega las herramientas en cualquier llamada API
 
 ## Qué Sucede Automáticamente en Claude Code
 
-No necesitas recordar todo manualmente. MeMesh tiene **6 hooks** que capturan e inyectan conocimiento mientras trabajas:
+No necesitas recordar todo manualmente. MeMesh tiene **7 hooks** que capturan e inyectan conocimiento mientras trabajas:
 
 | Cuándo | Qué hace MeMesh |
 |---|---|
 | **Al inicio de cada sesión** | Carga tus memorias más relevantes + advertencias proactivas de lecciones pasadas + banner de orquestación agentica |
 | **Antes de editar archivos** | Recupera memorias vinculadas al archivo o proyecto antes de que Claude escriba código |
 | **Antes de comandos bash** | Nudge a Claude para que envíe comandos de alta verificabilidad (test, build, lint, migrate, deploy, benchmark) como agentes de fondo |
+| **Cuando pides recordar** | Detecta intención de "remember this" / "記下來" y recuerda a Claude que escriba dual (memesh + MEMORY.md) |
 | **Después de cada `git commit`** | Registra qué cambiaste, con estadísticas de diff |
 | **Cuando Claude se detiene** | Captura archivos editados, errores corregidos y genera automáticamente lecciones estructuradas a partir de fallos |
 | **Antes de compresión de contexto** | Guarda conocimiento antes de que se pierda en límites de contexto |
@@ -271,7 +283,7 @@ El core es agnóstico de framework. La misma lógica se ejecuta desde terminal,
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.fr.md

@@ -37,6 +37,17 @@ Ce package constitue la couche de mémoire locale de la famille de produits MeMe
 npm install -g @pcircle/memesh
 ```
 
+### Étape 1,5 : Connecter MeMesh à Claude Code (recommandé, une seule fois)
+
+`npm install -g` place la CLI dans le PATH et enregistre le serveur MCP, mais **ne connecte pas** automatiquement les hooks de session MeMesh à Claude Code. Sans ces hooks, vous pouvez utiliser `memesh remember` / `recall` manuellement, mais la **boucle d'auto-capture** (session → leçons → rappel proactif à la session suivante) reste silencieuse.
+
+```bash
+memesh install-hooks         # ajoute les hooks memesh à ~/.claude/settings.json
+memesh doctor                # vérifie que « Hooks wired into Claude Code » passe
+```
+
+Ces hooks coexistent avec vos hooks personnalisés dans `~/.claude/hooks/` — `install-hooks` écrit de manière additive et n'écrase jamais les vôtres. Pour supprimer : `memesh uninstall-hooks`.
+
 ### Étape 2 : Mémoriser une décision
 
 ```bash
@@ -146,13 +157,14 @@ Collez les outils dans n'importe quel appel API
 
 ## Ce Qui Se Passe Automatiquement Dans Claude Code
 
-Vous n'avez pas besoin de tout mémoriser manuellement. MeMesh possède **6 hooks** qui capturent et injectent les connaissances au fur et à mesure que vous travaillez :
+Vous n'avez pas besoin de tout mémoriser manuellement. MeMesh possède **7 hooks** qui capturent et injectent les connaissances au fur et à mesure que vous travaillez :
 
 | Quand | Ce que MeMesh fait |
 |---|---|
 | **Au début de chaque session** | Charge vos mémoires les plus pertinentes + avertissements proactifs des leçons passées + banneau d'orchestration agentique |
 | **Avant d'éditer des fichiers** | Rappelle les mémoires liées au fichier ou au projet avant que Claude ne rédige du code |
 | **Avant les commandes bash** | Encourage Claude à dispatcher les commandes très vérifiables (test, build, lint, migrate, deploy, benchmark) en tant qu'agents de fond |
+| **Lorsque vous demandez de mémoriser** | Détecte l'intention "remember this" / "記下來" et rappelle à Claude d'écrire en double (memesh + MEMORY.md) |
 | **Après chaque `git commit`** | Enregistre ce que vous avez modifié, avec les statistiques de diff |
 | **Quand Claude s'arrête** | Capture les fichiers édités, les erreurs corrigées et génère automatiquement des leçons structurées à partir des défaillances |
 | **Avant la compaction de contexte** | Sauvegarde les connaissances avant qu'elles ne soient perdues aux limites de contexte |
@@ -271,7 +283,7 @@ Le cœur est agnostique du framework. La même logique s'exécute depuis le term
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.ja.md

@@ -37,6 +37,17 @@
 npm install -g @pcircle/memesh
 ```
 
+### ステップ 1.5: MeMesh を Claude Code に接続(推奨、一度きり)
+
+`npm install -g` は CLI を PATH に配置し MCP サーバーを登録しますが、MeMesh の Claude Code セッションフックは自動的にはワイヤリング**されません**。フックがないと `memesh remember` / `recall` は手動で使えますが、**自動キャプチャループ**(セッション → レッスン → 次のセッションで自発的にリコール)はサイレントになります。
+
+```bash
+memesh install-hooks         # ~/.claude/settings.json に memesh フックを追加
+memesh doctor                # "Hooks wired into Claude Code" が PASS になることを確認
+```
+
+これらのフックは既存の `~/.claude/hooks/` カスタムフックと共存します — `install-hooks` は追加方式で書き込み、既存のものを上書きしません。削除する場合: `memesh uninstall-hooks`。
+
 ### ステップ 2: 決定を記録
 
 ```bash
@@ -271,7 +282,7 @@ memesh  # ダッシュボード → Settings タブを開く
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.ko.md

@@ -37,6 +37,17 @@
 npm install -g @pcircle/memesh
 ```
 
+### 1.5단계: MeMesh를 Claude Code에 연결(권장, 일회성)
+
+`npm install -g`는 CLI를 PATH에 추가하고 MCP 서버를 등록하지만, MeMesh의 Claude Code 세션 훅을 자동으로 연결하지는 **않습니다**. 훅이 없으면 `memesh remember` / `recall`을 수동으로 사용할 수 있지만, **자동 캡처 루프**(세션 → 교훈 → 다음 세션에서 능동적 회상)는 작동하지 않습니다.
+
+```bash
+memesh install-hooks         # ~/.claude/settings.json에 memesh 훅 추가
+memesh doctor                # "Hooks wired into Claude Code" PASS 확인
+```
+
+이 훅들은 기존 `~/.claude/hooks/` 사용자 훅과 공존합니다 — `install-hooks`는 추가 방식으로 작성하며 기존 항목을 덮어쓰지 않습니다. 제거하려면: `memesh uninstall-hooks`.
+
 ### 2단계: 의사결정 저장
 
 ```bash
@@ -271,7 +282,7 @@ memesh  # 대시보드 열기 → Settings 탭
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.md

@@ -54,6 +54,17 @@ npm install -g @pcircle/memesh
 > - **Native modules** — `better-sqlite3` and `sqlite-vec` install via prebuilt binaries on macOS (arm64/x64), Linux (x64/arm64), and Windows x64. On uncommon platforms or when prebuilds fail, you'll need a working C/C++ toolchain.
 > - **Embedding model** — the first call that triggers a local embedding (e.g. `recall` with semantic mode) downloads `Xenova/all-MiniLM-L6-v2` (~80 MB) into `~/.memesh/models/`. Subsequent calls are instant. The default retrieval path (FTS5) does not require this download.
 
+### Step 1.5: Wire MeMesh into Claude Code (recommended, one-time)
+
+`npm install -g` puts the CLI on your PATH and registers the MCP server, but does **not** auto-wire MeMesh's Claude Code session hooks. Without these hooks, you can use `memesh remember` / `recall` manually but the **auto-capture loop** (sessions → lessons → recall on next session) is silent.
+
+```bash
+memesh install-hooks         # adds memesh's hooks to ~/.claude/settings.json
+memesh doctor                # verifies "Hooks wired into Claude Code" passes
+```
+
+The hooks coexist with any custom hooks you already have under `~/.claude/hooks/` — `install-hooks` writes additive entries and never overwrites yours. To remove later: `memesh uninstall-hooks`.
+
 ### Step 2: Store a decision
 
 ```bash
@@ -169,13 +180,14 @@ Paste tools into any API call
 
 ## What Happens Automatically In Claude Code
 
-You don't need to manually remember everything. MeMesh has **6 hooks** that capture and inject knowledge while you work:
+You don't need to manually remember everything. MeMesh has **7 hooks** that capture and inject knowledge while you work:
 
 | When | What MeMesh does |
 |------|------------------|
 | **Every session start** | Loads your most relevant memories + proactive warnings from past lessons |
 | **Before editing files** | Recalls memories tied to the file or project before Claude writes code |
 | **Before bash commands** | (Opt-in) Nudges Claude to dispatch high-verifiability commands (test, build, lint, migrate, deploy, benchmark) as background agents |
+| **When you ask to remember** | Detects "remember this" / "guardar en memesh" / "sauvegarder dans memesh" / "記下來" intent (5 languages) and reminds Claude to use memesh |
 | **After every `git commit`** | Records what you changed, with diff stats |
 | **When Claude stops** | Captures files edited, errors fixed, and auto-generates structured lessons from failures |
 | **Before context compaction** | Saves knowledge before it's lost to context limits |
@@ -194,13 +206,13 @@ All configuration is via environment variables. Defaults are local-only and zero
 | `MEMESH_AUTO_CAPTURE` | `true` | Disable the auto-capture hooks (`Stop`, `PreCompact`) entirely. |
 | `MEMESH_AUTO_DETECT_LLM` | unset | Set to `1` to let memesh auto-detect a provider from your shell env (`OPENAI_API_KEY` etc.) and switch to BYOK embeddings. **Default fresh-install is local ONNX (384-dim) only** — opt in if you want cloud embeddings. Without this flag set, an `OPENAI_API_KEY` lying around in your shell is ignored. |
 | `MEMESH_ENABLE_AGENTIC_ORCHESTRATION` | unset | Set to `1` to enable an experimental working-model protocol (CTO / Orchestrator / Agents framing). Adds a session-start banner, a Bash command nudge, and `verify_agent_work` telemetry. The protocol's effectiveness is being instrumented, not yet proven — opt in if you want to participate. **Default is OFF**: the core memory features work without this flag. |
-| `MEMESH_AUTO_UPDATE` | `off` | Auto-update policy applied at session-start. `off` (default) never auto-updates; `patch` allows `X.Y.Z → X.Y.Z+N`; `minor` adds `X.Y.Z → X.Y+1.0`; `major` allows any bump. The actual `npm install -g` runs detached so it never blocks the session — outcomes land in `~/.memesh/auto-update.log`. Also settable as `autoUpdate` in `~/.memesh/config.json` (env wins). When the installed version is deprecated by maintainers (security advisory), `patch` is force-allowed even on `off` — minor / major bumps still stay manual to avoid silent behaviour drift. |
+| `MEMESH_AUTO_UPDATE` | `off` | Auto-update policy. `off` (default) never auto-updates; `patch` allows `X.Y.Z → X.Y.Z+N`; `minor` adds `X.Y.Z → X.Y+1.0`; `major` allows any bump. When permitted, a detached `npm install -g` fires at session end (Stop hook) so it never blocks your work — outcomes land in `~/.memesh/auto-update.log`. Also settable as `autoUpdate` in `~/.memesh/config.json` (env wins). When the installed version is deprecated by maintainers (security advisory), `patch` is force-allowed even on `off` — minor / major bumps still stay manual to avoid silent behaviour drift. |
 | `OPENAI_API_KEY` | unset | Your OpenAI key. Only used when `MEMESH_AUTO_DETECT_LLM=1` or you explicitly configure the provider. |
 | `OLLAMA_HOST` | `http://localhost:11434` | Override the Ollama endpoint when using a local Ollama provider. |
 
 `memesh doctor` prints the resolved configuration so you can see what's active.
 
-When npm flags an installed version as deprecated (typically a security advisory), the next session-start prepends a strong `⚠️ MeMesh <ver> is DEPRECATED` banner and `memesh update-status` surfaces the same line until you upgrade. The check is cached at `~/.memesh/update-check.json` so a transient network failure can't dim the warning.
+When npm flags an installed version as deprecated (typically a security advisory), the next session-start prepends a strong `⚠️ MeMesh <ver> is DEPRECATED` banner and `memesh update-status` surfaces the same line until you upgrade. The check is cached at `~/.memesh/update-check.<version>.json` so a transient network failure can't dim the warning.
 
 ---
 
@@ -314,7 +326,7 @@ Core is framework-agnostic. Same logic runs from terminal, HTTP, or MCP.
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.pt.md

@@ -37,6 +37,17 @@ Este pacote é a camada de memória local da família de produtos MeMesh. É pro
 npm install -g @pcircle/memesh
 ```
 
+### Passo 1.5: Conecte o MeMesh ao Claude Code (recomendado, uma só vez)
+
+`npm install -g` coloca a CLI no PATH e registra o servidor MCP, mas **não** conecta automaticamente os hooks de sessão do MeMesh ao Claude Code. Sem esses hooks você pode usar `memesh remember` / `recall` manualmente, mas o **loop de auto-captura** (sessão → lições → recall proativo na próxima sessão) fica silencioso.
+
+```bash
+memesh install-hooks         # adiciona os hooks do memesh em ~/.claude/settings.json
+memesh doctor                # confirma que "Hooks wired into Claude Code" passou
+```
+
+Os hooks coexistem com qualquer hook customizado em `~/.claude/hooks/` — `install-hooks` escreve de forma aditiva e nunca sobrescreve. Para remover: `memesh uninstall-hooks`.
+
 ### Passo 2: Armazene uma decisão
 
 ```bash
@@ -146,13 +157,14 @@ Cole as ferramentas em qualquer chamada de API
 
 ## O Que Acontece Automaticamente no Claude Code
 
-Você não precisa lembrar tudo manualmente. MeMesh tem **6 hooks** que capturam e injetam conhecimento enquanto você trabalha:
+Você não precisa lembrar tudo manualmente. MeMesh tem **7 hooks** que capturam e injetam conhecimento enquanto você trabalha:
 
 | Quando | O que MeMesh faz |
 |------|------------------|
 | **Início de cada sessão** | Carrega suas memórias mais relevantes + alertas proativos de lições passadas + banner de orquestração de agentes |
 | **Antes de editar arquivos** | Recupera memórias vinculadas ao arquivo ou projeto antes de Claude escrever código |
 | **Antes de comandos bash** | Incentiva Claude a despachar comandos de alta verificabilidade (test, build, lint, migrate, deploy, benchmark) como agentes de background |
+| **Quando você pede para lembrar** | Detecta intenção "remember this" / "記下來" e lembra Claude de escrever dual (memesh + MEMORY.md) |
 | **Depois de cada `git commit`** | Registra o que você mudou, com estatísticas de diff |
 | **Quando Claude para** | Captura arquivos editados, erros corrigidos e gera automaticamente lições estruturadas de falhas |
 | **Antes da compactação de contexto** | Salva conhecimento antes de ser perdido nos limites de contexto |
@@ -271,7 +283,7 @@ Core é agnóstico a framework. A mesma lógica roda de terminal, HTTP ou MCP.
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.th.md

@@ -37,6 +37,17 @@
 npm install -g @pcircle/memesh
 ```
 
+### ขั้นตอนที่ 1.5: เชื่อม MeMesh เข้ากับ Claude Code (แนะนำ, ครั้งเดียว)
+
+`npm install -g` ติดตั้ง CLI ลงใน PATH และลงทะเบียน MCP server แต่**ไม่ได้**เชื่อม session hooks ของ MeMesh เข้ากับ Claude Code โดยอัตโนมัติ ถ้าไม่มี hooks เหล่านี้ คุณยังใช้ `memesh remember` / `recall` แบบ manual ได้ แต่**วงรอบจับข้อมูลอัตโนมัติ** (session → บทเรียน → เรียกคืนเชิงรุกใน session ถัดไป) จะเงียบ
+
+```bash
+memesh install-hooks         # เพิ่ม hooks ของ memesh ลงใน ~/.claude/settings.json
+memesh doctor                # ยืนยันว่า "Hooks wired into Claude Code" PASS
+```
+
+Hooks เหล่านี้อยู่ร่วมกับ custom hooks ที่คุณมีใน `~/.claude/hooks/` — `install-hooks` เขียนแบบเพิ่ม ไม่เขียนทับของคุณ หากต้องการลบ: `memesh uninstall-hooks`
+
 ### ขั้นตอนที่ 2: เก็บการตัดสินใจ
 
 ```bash
@@ -146,13 +157,14 @@ memesh export-schema \
 
 ## สิ่งที่เกิดขึ้นโดยอัตโนมัติใน Claude Code
 
-ไม่ต้องจำข้อมูลทุกอย่างด้วยตนเอง MeMesh มี **6 hook** ที่บันทึกและแทรกความรู้ขณะทำงาน:
+ไม่ต้องจำข้อมูลทุกอย่างด้วยตนเอง MeMesh มี **7 hook** ที่บันทึกและแทรกความรู้ขณะทำงาน:
 
 | เมื่อ | MeMesh ทำอะไร |
 |---|---|
 | **เริ่มเซสชันทุกครั้ง** | โหลดหน่วยความจำที่เกี่ยวข้องมากที่สุด + คำเตือนจากบทเรียนที่ผ่านมา + แบนเนอร์การ orchestrate เอเจนต์ |
 | **ก่อนแก้ไขไฟล์** | เรียกคืนหน่วยความจำที่เชื่อมโยงกับไฟล์หรือโครงการก่อนที่ Claude เขียนโค้ด |
 | **ก่อนคำสั่ง bash** | แนะนำให้ Claude สั่งคำสั่งที่ยืนยันได้สูง (test build lint migrate deploy benchmark) เป็นเอเจนต์พื้นหลัง |
+| **เมื่อคุณขอให้จำ** | ตรวจจับความตั้งใจ "remember this" / "記下來" และเตือน Claude ให้เขียนแบบสองทาง (memesh + MEMORY.md) |
 | **หลังทุก `git commit`** | บันทึกสิ่งที่เปลี่ยนแปลง พร้อมสถิติ diff |
 | **เมื่อ Claude หยุด** | บันทึกไฟล์ที่แก้ไข บั๊กที่แก้ไข และสร้างบทเรียนโครงสร้างจากความล้มเหลวโดยอัตโนมัติ |
 | **ก่อนการบีบอัด context** | บันทึกความรู้ก่อนที่จะหายไปจากขีดจำกัด context |
@@ -271,7 +283,7 @@ Core เป็นอิสระจากเฟรมเวิร์ก ตร
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.vi.md

@@ -37,6 +37,17 @@ Package này là tầng bộ nhớ cục bộ của dòng sản phẩm MeMesh. N
 npm install -g @pcircle/memesh
 ```
 
+### Bước 1.5: Kết nối MeMesh vào Claude Code (khuyến nghị, một lần)
+
+`npm install -g` đặt CLI vào PATH và đăng ký MCP server, nhưng **không** tự động kết nối các session hooks của MeMesh vào Claude Code. Không có hooks, bạn vẫn dùng được `memesh remember` / `recall` thủ công, nhưng **vòng tự động ghi nhận** (session → bài học → gọi lại chủ động ở session sau) sẽ im lặng.
+
+```bash
+memesh install-hooks         # thêm hooks memesh vào ~/.claude/settings.json
+memesh doctor                # xác nhận "Hooks wired into Claude Code" PASS
+```
+
+Các hooks này cùng tồn tại với bất kỳ hook tùy chỉnh nào trong `~/.claude/hooks/` — `install-hooks` ghi theo kiểu thêm, không bao giờ ghi đè của bạn. Để gỡ: `memesh uninstall-hooks`.
+
 ### Bước 2: Lưu một quyết định
 
 ```bash
@@ -146,13 +157,14 @@ Dán tools vào bất kỳ API call nào
 
 ## Điều gì xảy ra tự động trong Claude Code
 
-Bạn không cần phải manually nhớ mọi thứ. MeMesh có **6 hooks** để capture và inject kiến thức khi bạn làm việc:
+Bạn không cần phải manually nhớ mọi thứ. MeMesh có **7 hooks** để capture và inject kiến thức khi bạn làm việc:
 
 | Khi nào | MeMesh làm gì |
 |------|------------------|
 | **Mỗi lần session bắt đầu** | Load những memories liên quan nhất + cảnh báo chủ động từ bài học trong quá khứ + agentic-orchestration banner |
 | **Trước khi chỉnh sửa file** | Gọi lại memories liên quan đến file hoặc dự án trước khi Claude viết code |
 | **Trước bash commands** | Hướng dẫn Claude dispatch những commands có độ xác minh cao (test, build, lint, migrate, deploy, benchmark) dưới dạng background agents |
+| **Khi bạn yêu cầu ghi nhớ** | Phát hiện ý định "remember this" / "記下來" và nhắc nhở (use memesh|ghi memesh) |
 | **Sau mỗi `git commit`** | Ghi lại những gì bạn thay đổi, với diff stats |
 | **Khi Claude dừng** | Capture những file đã chỉnh sửa, lỗi đã sửa, và auto-generate structured lessons từ failures |
 | **Trước khi context compact** | Lưu kiến thức trước khi nó bị mất do context limits |
@@ -271,7 +283,7 @@ Core là framework-agnostic. Logic tương tự chạy từ terminal, HTTP, ho
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 tests
+npm test             # 630 tests
 npm run test:e2e-dashboard
 ```
 

package/README.zh-CN.md

@@ -37,6 +37,17 @@
 npm install -g @pcircle/memesh
 ```
 
+### 第一步半：把 MeMesh 接入 Claude Code（推荐，一次性）
+
+`npm install -g` 把 CLI 放进 PATH 并注册 MCP server，但**不会**自动接上 MeMesh 的 Claude Code session hooks。没有这些 hooks，你还是可以手动用 `memesh remember` / `recall`，但**自动捕捉循环**（session → 教训 → 下次 session 主动回想）就会静默不动。
+
+```bash
+memesh install-hooks         # 把 memesh hooks 加到 ~/.claude/settings.json
+memesh doctor                # 确认「Hooks wired into Claude Code」通过
+```
+
+这些 hooks 会和你已有的 `~/.claude/hooks/` 自定义 hooks 并存 — `install-hooks` 用追加方式写，从不覆盖你的东西。要移除：`memesh uninstall-hooks`。
+
 ### 第二步：记录一个决策
 
 ```bash
@@ -271,7 +282,7 @@ memesh  # 打开仪表板 → 设置标签页
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 个测试
+npm test             # 630 个测试
 npm run test:e2e-dashboard
 ```
 

package/README.zh-TW.md

@@ -37,6 +37,17 @@
 npm install -g @pcircle/memesh
 ```
 
+### 第一步半：把 MeMesh 接進 Claude Code（建議，一次性）
+
+`npm install -g` 把 CLI 放上 PATH 並註冊 MCP server，但**不會**自動接上 MeMesh 的 Claude Code session hooks。沒有這些 hooks，你還是可以手動用 `memesh remember` / `recall`，但**自動捕捉迴路**（session → 教訓 → 下次 session 主動回憶）就會靜默不動。
+
+```bash
+memesh install-hooks         # 把 memesh hooks 加進 ~/.claude/settings.json
+memesh doctor                # 確認「Hooks wired into Claude Code」過了
+```
+
+這些 hooks 會跟你既有的 `~/.claude/hooks/` 自訂 hooks 共存 — `install-hooks` 用追加方式寫，從不覆寫你的東西。要移除：`memesh uninstall-hooks`。
+
 ### 第二步：保存一個決策
 
 ```bash
@@ -271,7 +282,7 @@ memesh  # 開啟儀表板 → 設定標籤
 ```bash
 git clone https://github.com/PCIRCLE-AI/memesh-llm-memory
 cd memesh-llm-memory && npm install && npm run build
-npm test             # 489 項測試
+npm test             # 630 項測試
 npm run test:e2e-dashboard
 ```