@shahmilsaari/memory-core 0.1.2 → 0.1.4

package/README.md

@@ -1,75 +1,46 @@
-# @shahmilsaari/memory-core
+# memory-core
 
-Universal AI memory for any project. Install once, and every AI coding agent — Copilot, Cursor, Claude Code, Windsurf, Cline, and 13 more — automatically follows your architecture rules, past decisions, and best practices. Backed by PostgreSQL + pgvector and fully local AI via Ollama. No OpenAI key required.
+**Every AI coding agent in your project, following your rules. Automatically.**
+
+You decide the architecture. memory-core remembers it. Every AI tool — Copilot, Cursor, Claude Code, Windsurf, and 10 more — reads those rules before writing a single line of code.
 
 ```bash
 npx @shahmilsaari/memory-core init
 ```
 
----
-
-## How it works
-
-```
-memory-core init
-      │
-      ├── You pick: Backend / Frontend / Fullstack
-      ├── You pick: Architecture (Clean, MVC, React, Vue, …)
-      │
-      ▼
-PostgreSQL + pgvector
-192 predefined rules — each with a WHY
-+ your own saved decisions
-      │
-      ▼
-18 AI agent config files generated
-      │
-      ▼
-Every agent reads its file → follows your rules
-
-git commit → pre-commit hook → Ollama checks diff
-→ violations blocked with Rule + Why + Fix
-```
+No OpenAI key. No cloud. Fully local.
 
 ---
 
-## Supported agents
+## What does it actually do?
 
-| Agent | File generated |
-|---|---|
-| Claude Code | `CLAUDE.md` |
-| GitHub Copilot | `.github/copilot-instructions.md` |
-| Cursor | `.cursorrules` + `.cursor/rules/memory-core.mdc` |
-| Windsurf | `.windsurfrules` |
-| Cline | `.clinerules` |
-| Roo Code | `.roo/rules/memory-core.md` |
-| Aider | `.aider.conf.yml` |
-| Continue.dev | `.continue/config.json` |
-| Devin | `DEVIN.md` |
-| Amazon Q | `.amazonq/dev/guidelines.md` |
-| Gemini Code Assist | `.gemini/styleguide.md` |
-| Zed AI | `.zed/settings.json` |
-| JetBrains AI | `.idea/ai-instructions.md` |
-| OpenHands / AGENTS | `AGENTS.md` |
+Without memory-core, every AI agent starts fresh. It doesn't know you're using Clean Architecture. It doesn't know controllers shouldn't call the database. It doesn't know why you made that decision six months ago.
 
-Plus: `AI_RULES.md`, `ARCHITECTURE.md`, `PROJECT_MEMORY.md`
+With memory-core:
+
+1. You run `init` once and answer a few questions about your project
+2. memory-core generates config files for every AI agent you use
+3. Those agents read the files and follow your rules — automatically
+4. When you commit code, a hook catches violations before they land in the repo
+5. Watch mode catches violations as you type, not just at commit time
 
 ---
 
-## Prerequisites
+## Before you start
 
-| Requirement | Purpose |
-|---|---|
-| PostgreSQL 14+ | Memory storage |
-| pgvector | Semantic similarity search |
-| Ollama + `nomic-embed-text` | Free local embeddings (no API key) |
-| Ollama + `llama3.2` | Pre-commit rule checking |
+You need three things installed:
+
+| What | Why | Install |
+|---|---|---|
+| PostgreSQL 14+ | Stores your rules and decisions | See below |
+| pgvector | Makes search smart (finds similar rules, not just exact matches) | See below |
+| Ollama | Runs AI locally — no API key needed | See below |
 
 ---
 
-## Setup
+## Install
 
-### 1. PostgreSQL
+### PostgreSQL
 
 **macOS**
 ```bash
@@ -83,13 +54,20 @@ sudo apt install postgresql postgresql-contrib
 sudo systemctl start postgresql
 ```
 
-**Windows** — download from [postgresql.org](https://www.postgresql.org/download/windows/)
+**Windows** — [download from postgresql.org](https://www.postgresql.org/download/windows/)
 
 ---
 
-### 2. pgvector
+### pgvector
+
+pgvector adds AI-powered search to PostgreSQL.
+
+**macOS + PostgreSQL 17+** (easiest)
+```bash
+brew install pgvector
+```
 
-**macOS with PostgreSQL@16** (brew pgvector only targets pg17+, build from source):
+**macOS + PostgreSQL 16** (brew's pgvector targets 17+, so build from source)
 ```bash
 xcode-select --install
 git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git /tmp/pgvector
@@ -98,24 +76,21 @@ make PG_CONFIG=/opt/homebrew/opt/postgresql@16/bin/pg_config
 make install PG_CONFIG=/opt/homebrew/opt/postgresql@16/bin/pg_config
 ```
 
-**macOS with PostgreSQL@17+**
-```bash
-brew install pgvector
-```
-
 **Linux**
 ```bash
 sudo apt install postgresql-16-pgvector   # replace 16 with your version
 ```
 
-Verify:
+Check it worked:
 ```bash
 psql -U $(whoami) -c "SELECT * FROM pg_available_extensions WHERE name = 'vector';"
 ```
 
 ---
 
-### 3. Ollama
+### Ollama
+
+Ollama runs AI models on your machine. Two models are needed — one for search, one for code checking.
 
 **macOS**
 ```bash
@@ -129,28 +104,24 @@ curl -fsSL https://ollama.com/install.sh | sh
 ollama serve &
 ```
 
-**Windows** — download from [ollama.com](https://ollama.com/download)
+**Windows** — [download from ollama.com](https://ollama.com/download)
 
-Pull the required models:
+Then pull the two models:
 ```bash
-ollama pull nomic-embed-text   # embeddings — used by remember/search/sync
-ollama pull llama3.2           # code analysis — used by pre-commit hook
+ollama pull nomic-embed-text   # used for search
+ollama pull llama3.2           # used to check your code
 ```
 
 ---
 
-### 4. Create the database
+### Create the database
 
 ```bash
 createdb memory_core
-```
-
-Then run the schema (get `setup.sql` from the repo or copy below):
-```bash
 psql -U $(whoami) -d memory_core -f setup.sql
 ```
 
-`setup.sql` contents:
+`setup.sql`:
 ```sql
 CREATE EXTENSION IF NOT EXISTS vector;
 
@@ -169,38 +140,31 @@ CREATE TABLE IF NOT EXISTS memories (
 );
 
 CREATE INDEX IF NOT EXISTS memories_embedding_idx
-  ON memories USING ivfflat (embedding vector_cosine_ops)
-  WITH (lists = 100);
-
+  ON memories USING ivfflat (embedding vector_cosine_ops) WITH (lists = 100);
 CREATE INDEX IF NOT EXISTS memories_architecture_idx ON memories (architecture);
 CREATE INDEX IF NOT EXISTS memories_scope_idx ON memories (scope);
 ```
 
 ---
 
-### 5. Configure environment
-
-Create `.env` in the project where you run memory-core (or `.memory-core.env` per project):
-
-```env
-DATABASE_URL=postgresql://YOUR_USERNAME@localhost:5432/memory_core
-OLLAMA_URL=http://localhost:11434
-OLLAMA_MODEL=nomic-embed-text
-OLLAMA_CHAT_MODEL=llama3.2
-```
-
-Find your username: `whoami`
-
----
+## Quick start
 
-### 6. Seed the database
+```bash
+# 1. Go to your project
+cd my-api
 
-Load 192 predefined best-practice rules (each with a reason explaining why it exists):
+# 2. Initialize — answers a few questions, generates all config files
+npx @shahmilsaari/memory-core init
 
-```bash
+# 3. Load 192 predefined best-practice rules
 npx @shahmilsaari/memory-core seed
+
+# 4. Install the pre-commit hook (optional but recommended)
+npx @shahmilsaari/memory-core hook install
 ```
 
+That's it. Every AI agent in your project now has your rules.
+
 ---
 
 ## Commands
@@ -208,107 +172,108 @@ npx @shahmilsaari/memory-core seed
 ### `init` — Set up a project
 
 ```bash
-cd your-project
 npx @shahmilsaari/memory-core init
 ```
 
-Interactive prompts:
-```
-Project name?         → my-api
-Project type?         → Backend / Frontend / Fullstack
-Backend architecture? → Clean Architecture
-Frontend framework?   → (skipped for backend-only)
-Language?             → TypeScript
-Install caveman?      → no
-```
-
-Generates all 18 agent config files and saves `.memory-core.json`.
-
----
-
-### `sync` — Regenerate all agent files
-
-```bash
-npx @shahmilsaari/memory-core sync
-```
+Asks you:
+- Project name
+- Project type (Backend / Frontend / Fullstack)
+- Architecture (Clean Architecture, MVC, React, Vue, etc.)
+- Language
+- Whether to enable caveman mode (optional token saver)
 
-Re-pulls memories from the DB and rewrites all 18 files. Run after `remember` or `seed`.
+Generates config files for every supported AI agent and saves your choices to `.memory-core.json`.
 
 ---
 
 ### `remember` — Save a decision
 
+Made a decision your team should never forget? Save it.
+
 ```bash
 npx @shahmilsaari/memory-core remember "Controllers must never call the database directly"
 ```
 
-Prompts: _"Why does this rule exist?"_ — the reason is stored and shown to agents and developers when a violation is caught.
+It will ask you *why* — that reason gets stored alongside the rule and shown to AI agents and developers when a violation is caught.
 
-With flags:
+With flags (skip the prompts):
 ```bash
 npx @shahmilsaari/memory-core remember "Use DTOs for all API responses" \
   --type rule \
   --scope global \
-  --reason "Raw DB entities leak internal schema and expose sensitive fields to clients" \
-  --tags "api,dto,response"
+  --reason "Raw DB entities expose internal schema and sensitive fields" \
+  --tags "api,dto"
 ```
 
-| Flag | Values | Default |
+| Flag | Options | Default |
 |---|---|---|
 | `--type` | `decision` `rule` `pattern` `note` | `decision` |
 | `--scope` | `global` `project` | `project` |
-| `--reason` | free text | prompted interactively |
+| `--reason` | any text | asked interactively |
 | `--tags` | comma-separated | none |
 
 ---
 
-### `search` — Semantic search
+### `sync` — Refresh all agent files
+
+After saving new memories, regenerate every agent file to include them.
+
+```bash
+npx @shahmilsaari/memory-core sync
+```
+
+---
+
+### `search` — Find a rule or decision
 
 ```bash
 npx @shahmilsaari/memory-core search "error handling"
 npx @shahmilsaari/memory-core search "auth strategy" --limit 10
 ```
 
+Uses AI search — finds related rules even if you don't use the exact words.
+
 ---
 
 ### `seed` — Load predefined rules
 
+192 best-practice rules across all supported architectures, each with a plain-English reason explaining why the rule exists.
+
 ```bash
-npx @shahmilsaari/memory-core seed
-npx @shahmilsaari/memory-core seed --arch clean-architecture   # one architecture only
-npx @shahmilsaari/memory-core seed --force                     # re-seed existing entries
+npx @shahmilsaari/memory-core seed                              # all architectures
+npx @shahmilsaari/memory-core seed --arch clean-architecture   # one only
+npx @shahmilsaari/memory-core seed --force                     # re-seed existing
 ```
 
-192 rules across all architectures. Every rule includes a `reason` explaining why it exists.
-
 ---
 
-### `hook install` — Pre-commit enforcement
+### `hook install` — Block bad commits
 
 ```bash
 npx @shahmilsaari/memory-core hook install
 ```
 
-Installs a git pre-commit hook. Every `git commit` automatically checks staged source files against your architecture rules using Ollama (llama3.2). Violations are blocked with full context:
+Installs a git pre-commit hook. Every time you run `git commit`, your staged files are checked against your architecture rules before the commit goes through.
+
+When a violation is found, the commit is blocked and you see exactly what's wrong and how to fix it:
 
 ```
   ✗ 2 rule violations found — commit blocked
 
   [1] src/controllers/user.ts:32
       Rule:   Thin controllers — business logic belongs in services
-      Why:    Logic in controllers cannot be reused from CLI, queues, or other
-              entry points — it gets siloed and duplicated across handlers
+      Why:    Logic in controllers cannot be reused from other entry points
+              — it gets siloed and duplicated across handlers
       Issue:  Password validation logic inside route handler
       Fix:    Move to UserService.validateCredentials()
 
   [2] src/domain/user.entity.ts:5
       Rule:   Domain has zero external imports
-      Why:    Framework imports in the domain layer tie business logic to
-              infrastructure — you can't swap the DB or test without spinning up the stack
+      Why:    Framework imports tie business logic to infrastructure
       Issue:  Imports 'typeorm' directly
-      Fix:    Define IUserRepository interface in domain, implement in infrastructure/
+      Fix:    Define IUserRepository interface in domain/
 
-  To bypass (not recommended): git commit --no-verify
+  To bypass: git commit --no-verify
   To save as memory: memory-core remember "<lesson>"
 ```
 
@@ -318,26 +283,63 @@ npx @shahmilsaari/memory-core hook uninstall   # remove the hook
 
 ---
 
-### `check` — Manual rule check
+### `watch` — Catch violations as you type
 
 ```bash
-npx @shahmilsaari/memory-core check --staged           # check staged files
-npx @shahmilsaari/memory-core check --staged --verbose # show model + diff details
+npx @shahmilsaari/memory-core watch
+```
+
+Runs in the background and checks each file the moment you save it. You see violations immediately — before you even think about committing.
+
 ```
+  archmind watch — real-time rule enforcement
 
-Same as the pre-commit hook — useful in CI/CD pipelines.
+  watching: /your/project
+  model:    llama3.2
+  rules:    24
+  ctrl+c to stop
+
+  [10:42:11] saved: src/controllers/user.ts
+
+  ✗ 1 violation in src/controllers/user.ts
+
+  [1] src/controllers/user.ts:34
+      Rule:   Thin controllers — business logic belongs in services
+      Why:    Logic in controllers cannot be reused from other entry points
+      Issue:  Password hashing inside the route handler
+      Fix:    Move to UserService.hashPassword()
+```
+
+Options:
+```bash
+npx @shahmilsaari/memory-core watch --path src/   # watch a specific folder only
+npx @shahmilsaari/memory-core watch --verbose     # show extra details
+```
+
+Only checks source files — ignores `node_modules`, `dist`, config files, JSON, etc.
 
 ---
 
-### `global` — Sync to all agents globally
+### `check` — Manual check (for CI)
 
-Writes your rules to the global config of every installed agent — so they follow your rules in every project without running `init`:
+```bash
+npx @shahmilsaari/memory-core check --staged           # check staged files
+npx @shahmilsaari/memory-core check --staged --verbose # with extra detail
+```
+
+Same as the pre-commit hook. Use this in CI/CD pipelines.
+
+---
+
+### `global` — Apply rules to every project
 
 ```bash
 npx @shahmilsaari/memory-core global
 ```
 
-| Agent | Global file |
+Writes your rules to the global config of each AI agent — so they follow your rules in every project on your machine, not just the current one.
+
+| Agent | Where it writes |
 |---|---|
 | Claude Code | `~/.claude/CLAUDE.md` |
 | GitHub Copilot | VS Code `settings.json` |
@@ -350,63 +352,89 @@ npx @shahmilsaari/memory-core global
 
 ---
 
+## Supported agents
+
+| Agent | File generated |
+|---|---|
+| Claude Code | `CLAUDE.md` |
+| GitHub Copilot | `.github/copilot-instructions.md` |
+| Cursor | `.cursorrules` + `.cursor/rules/memory-core.mdc` |
+| Windsurf | `.windsurfrules` |
+| Cline | `.clinerules` |
+| Roo Code | `.roo/rules/memory-core.md` |
+| Aider | `.aider.conf.yml` |
+| Continue.dev | `.continue/config.json` |
+| Devin | `DEVIN.md` |
+| Amazon Q | `.amazonq/dev/guidelines.md` |
+| Gemini Code Assist | `.gemini/styleguide.md` |
+| Zed AI | `.zed/settings.json` |
+| JetBrains AI | `.idea/ai-instructions.md` |
+| OpenHands | `AGENTS.md` |
+
+Plus shared files: `AI_RULES.md`, `ARCHITECTURE.md`, `PROJECT_MEMORY.md`
+
+---
+
 ## Architecture profiles
 
-### Backend
-| Profile | Best for |
+Pick the one that matches how your project is structured.
+
+**Backend**
+| Profile | Use when… |
 |---|---|
-| **Clean Architecture** | Domain-driven APIs, strict layer separation |
-| **Modular Monolith** | Feature modules that could become microservices |
-| **MVC** | Standard web apps, controllers + services |
-| **Hexagonal** | Port/adapter isolation, highly testable cores |
-| **Next.js** | Next.js 13+ fullstack (server components + actions) |
-| **Laravel** | Laravel APIs with service-repository pattern |
-
-### Frontend
-| Profile | Best for |
+| Clean Architecture | You want strict separation between domain, application, and infrastructure |
+| Modular Monolith | You're building feature modules that might become microservices later |
+| MVC | Standard web app with controllers and services |
+| Hexagonal | You want ports and adapters for maximum testability |
+| Next.js | Next.js 13+ with server components and server actions |
+| Laravel | Laravel with service-repository pattern |
+
+**Frontend**
+| Profile | Use when… |
 |---|---|
-| **React** | Functional components, hooks, React Query, Zustand |
-| **Vue 3** | Composition API, Pinia, composables |
-| **Angular** | Standalone components, signals, OnPush |
-| **Svelte** | Svelte 5 runes, SvelteKit load functions |
-| **Nuxt 3** | Fullstack Vue with SSR, server routes |
-| **React Native** | Mobile apps, FlatList, secure storage |
+| React | Functional components, hooks, React Query, Zustand |
+| Vue 3 | Composition API, Pinia, composables |
+| Angular | Standalone components, signals, OnPush strategy |
+| Svelte | Svelte 5 runes, SvelteKit load functions |
+| Nuxt 3 | Fullstack Vue with SSR |
+| React Native | Mobile apps |
 
-Fullstack projects get dual sections — one for backend, one for frontend — in every generated agent file.
+Fullstack projects get both sections in every generated file.
 
 ---
 
-## Caveman token saver (optional)
+## Caveman mode (optional)
 
-Compresses AI responses 65–75% by stripping filler words. Enabled optionally during `init`. Injected into every agent file automatically.
+Cuts AI response length by 65–75% by removing filler words. Opt in during `init`.
 
-| Level | Style |
+| Level | What it does |
 |---|---|
-| `lite` | Professional terseness |
-| `full` | Caveman mode |
-| `ultra` | Telegraphic, minimum words |
+| `lite` | Professional and concise |
+| `full` | Caveman-style short answers |
+| `ultra` | Absolute minimum words |
 
 ---
 
-## Typical workflow
+## Day-to-day workflow
 
 ```bash
-# New project
+# Starting a new project
 cd my-api
 npx @shahmilsaari/memory-core init
+npx @shahmilsaari/memory-core seed
 npx @shahmilsaari/memory-core hook install
 
-# Save a decision mid-project
-npx @shahmilsaari/memory-core remember "All auth goes through middleware, never duplicated in controllers" \
+# Made an architectural decision? Save it.
+npx @shahmilsaari/memory-core remember "All auth goes through middleware, never in controllers" \
   --type decision --scope global
 
-# Sync agent files after new decisions
+# Refresh agent files after saving new memories
 npx @shahmilsaari/memory-core sync
 
-# Search before making an architectural decision
+# Not sure how something was decided? Search.
 npx @shahmilsaari/memory-core search "caching strategy"
 
-# Try to commit code that breaks rules → hook blocks it with explanation
+# Commit code → hook checks it automatically
 git commit -m "add user endpoint"
 ```
 
@@ -414,49 +442,51 @@ git commit -m "add user endpoint"
 
 ## Environment variables
 
-| Variable | Required | Default | Description |
+memory-core creates `.memory-core.env` automatically during `init`. You can also set these manually:
+
+| Variable | Required | Default | What it does |
 |---|---|---|---|
 | `DATABASE_URL` | Yes | — | PostgreSQL connection string |
-| `OLLAMA_URL` | No | `http://localhost:11434` | Ollama server URL |
-| `OLLAMA_MODEL` | No | `nomic-embed-text` | Embedding model |
-| `OLLAMA_CHAT_MODEL` | No | `llama3.2` | Chat model for pre-commit checks |
+| `OLLAMA_URL` | No | `http://localhost:11434` | Where Ollama is running |
+| `OLLAMA_MODEL` | No | `nomic-embed-text` | Model used for search |
+| `OLLAMA_CHAT_MODEL` | No | `llama3.2` | Model used for code checking |
 
 ---
 
 ## Troubleshooting
 
 **`extension "vector" is not available`**
-pgvector not installed for your PostgreSQL version. See [pgvector setup](#2-pgvector-setup).
+pgvector isn't installed for your PostgreSQL version. Follow the [pgvector install steps](#pgvector) above.
 
 **`Ollama not running — skipping rule check`**
 Start Ollama: `brew services start ollama` (macOS) or `ollama serve` (Linux).
 
 **`Chat model "llama3.2" not found`**
-Pull the model: `ollama pull llama3.2`. Or set `OLLAMA_CHAT_MODEL=mistral` in `.env`.
+Run `ollama pull llama3.2`. Or switch to another model: add `OLLAMA_CHAT_MODEL=mistral` to `.memory-core.env`.
 
 **`DATABASE_URL is not set`**
-Create `.env` in your project root. See [Configure environment](#5-configure-environment).
+Run `npx @shahmilsaari/memory-core init` — it will create the `.memory-core.env` file for you.
 
 **`createdb: role does not exist`**
 ```bash
 createuser -s $(whoami)
 ```
 
-**Pre-commit hook flagging config/JSON files**
-The hook only checks source files (`.ts .tsx .js .jsx .py .php .rb .go .java .cs .swift .kt .rs .vue .svelte`). Config and env files are automatically skipped.
+**Hook is flagging JSON or config files**
+It won't — the hook only checks source code files: `.ts .tsx .js .jsx .py .php .rb .go .java .cs .swift .kt .rs .vue .svelte`. Everything else is skipped automatically.
 
 ---
 
 ## Roadmap
 
-| Feature | Description |
+| | Feature |
 |---|---|
-| CI/CD check | GitHub Actions workflow — fails PRs that violate rules |
-| Watch mode | `memory-core watch` — checks files on save in real-time |
-| Violation memory | Auto-save caught violations as "never do X, do Y" memories |
-| Rule analytics | Track which rules break most, which files are worst offenders |
-| Team sync | `memory-core push/pull` — shared memory pool across the whole team |
-| Memory review | `memory-core approve` — team sign-off before rules propagate |
+| ✓ | Watch mode — real-time violation alerts on save |
+| | CI/CD — fail PRs that violate your rules |
+| | Violation memory — auto-save what went wrong as a new rule |
+| | Rule analytics — see which rules get broken most |
+| | Team sync — share memory across your whole team |
+| | Memory review — approve rules before they propagate |
 
 ---
 

package/dist/cli.js

@@ -3,12 +3,12 @@
 // src/cli.ts
 import { Command } from "commander";
 import { input, select, confirm } from "@inquirer/prompts";
-import chalk2 from "chalk";
+import chalk3 from "chalk";
 import ora from "ora";
-import { readFileSync as readFileSync4, writeFileSync as writeFileSync3, existsSync as existsSync5, mkdirSync as mkdirSync2, appendFileSync } from "fs";
-import { join as join5, dirname as dirname2 } from "path";
+import { readFileSync as readFileSync5, writeFileSync as writeFileSync3, existsSync as existsSync6, mkdirSync as mkdirSync2, appendFileSync } from "fs";
+import { join as join6, dirname as dirname2 } from "path";
 import { homedir } from "os";
-import { execSync as execSync2 } from "child_process";
+import { execSync as execSync3 } from "child_process";
 
 // src/generator.ts
 import { readFileSync, readdirSync, writeFileSync, mkdirSync, existsSync } from "fs";
@@ -626,10 +626,10 @@ function uninstallHook() {
   console.log(chalk.green("\n  \u2713 Pre-commit hook removed\n"));
 }
 async function checkStaged(options = {}) {
-  const SOURCE_EXTENSIONS = /\.(ts|tsx|js|jsx|py|php|rb|go|java|cs|swift|kt|rs|vue|svelte)$/;
+  const SOURCE_EXTENSIONS2 = /\.(ts|tsx|js|jsx|py|php|rb|go|java|cs|swift|kt|rs|vue|svelte)$/;
   let diff;
   try {
-    const stagedFiles = execSync("git diff --cached --name-only", { encoding: "utf-8" }).split("\n").filter((f) => f && SOURCE_EXTENSIONS.test(f));
+    const stagedFiles = execSync("git diff --cached --name-only", { encoding: "utf-8" }).split("\n").filter((f) => f && SOURCE_EXTENSIONS2.test(f));
     if (stagedFiles.length === 0) {
       if (options.verbose) console.log(chalk.gray("  No source files staged \u2014 skipping rule check."));
       return;
@@ -782,57 +782,283 @@ function printModelMissing(model) {
   console.log(chalk.gray("     Recommended: llama3.2  |  qwen2.5-coder:3b  |  mistral\n"));
 }
 
+// src/watcher.ts
+import { watch } from "chokidar";
+import { execSync as execSync2 } from "child_process";
+import { existsSync as existsSync5, readFileSync as readFileSync4 } from "fs";
+import { join as join5, relative } from "path";
+import chalk2 from "chalk";
+var SOURCE_EXTENSIONS = /\.(ts|tsx|js|jsx|py|php|rb|go|java|cs|swift|kt|rs|vue|svelte)$/;
+var reasonMap2 = new Map(
+  seeds.filter((s) => s.reason).map((s) => [s.content, s.reason])
+);
+function loadConfig(cwd) {
+  const configPath = join5(cwd, ".memory-core.json");
+  if (!existsSync5(configPath)) return null;
+  try {
+    return JSON.parse(readFileSync4(configPath, "utf-8"));
+  } catch {
+    return null;
+  }
+}
+function getProfileRules(config2) {
+  const rules = [];
+  const avoids = [];
+  if (config2.backendArchitecture) {
+    const profile = listProfiles("backend").find((p) => p.name === config2.backendArchitecture);
+    if (profile) {
+      rules.push(...profile.rules);
+      avoids.push(...profile.avoid);
+    }
+  }
+  if (config2.frontendFramework) {
+    const profile = listProfiles("frontend").find((p) => p.name === config2.frontendFramework);
+    if (profile) {
+      rules.push(...profile.rules);
+      avoids.push(...profile.avoid);
+    }
+  }
+  return { rules, avoids };
+}
+async function checkFile(filePath, cwd, config2, verbose) {
+  const rel = relative(cwd, filePath);
+  let diff;
+  try {
+    diff = execSync2(`git diff HEAD -- "${rel}" 2>/dev/null || git diff --no-index /dev/null "${rel}" 2>/dev/null || true`, {
+      encoding: "utf-8",
+      cwd
+    });
+  } catch {
+    try {
+      diff = execSync2(`git diff --no-index /dev/null "${rel}"`, { encoding: "utf-8", cwd });
+    } catch (e) {
+      diff = e.stdout ?? "";
+    }
+  }
+  if (!diff.trim()) return;
+  const { rules, avoids } = getProfileRules(config2);
+  if (rules.length === 0) return;
+  const ollamaUrl = process.env.OLLAMA_URL ?? "http://localhost:11434";
+  const chatModel = process.env.OLLAMA_CHAT_MODEL ?? "llama3.2";
+  const MAX_DIFF = 6e3;
+  const truncated = diff.length > MAX_DIFF;
+  const diffToSend = truncated ? diff.slice(0, MAX_DIFF) + "\n\n[diff truncated]" : diff;
+  if (verbose) {
+    console.log(chalk2.dim(`
+  [watch] checking ${rel} (${diff.length} chars)\u2026`));
+  }
+  const rulesWithReasons = rules.map((r, i) => {
+    const why = reasonMap2.get(r);
+    return why ? `${i + 1}. ${r}
+   WHY: ${why}` : `${i + 1}. ${r}`;
+  }).join("\n");
+  const systemPrompt = `You are a strict code reviewer enforcing architecture rules.
+Analyze the file diff and identify ONLY clear, definite rule violations.
+Use the WHY for each rule to understand intent and judge edge cases.
+
+Rules to enforce:
+${rulesWithReasons}
+
+Things that must never appear:
+${avoids.map((a, i) => `${i + 1}. ${a}`).join("\n")}
+
+IMPORTANT: Respond with JSON: {"violations":[...]} or {"violations":[]}.
+Each violation: {"rule":"...","file":"...","line":N,"issue":"...","suggestion":"...","reason":"..."}.
+No text outside the JSON.`;
+  try {
+    const res = await fetch(`${ollamaUrl}/api/chat`, {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({
+        model: chatModel,
+        messages: [
+          { role: "system", content: systemPrompt },
+          { role: "user", content: `Review this diff for ${rel}:
+
+${diffToSend}` }
+        ],
+        stream: false,
+        format: "json"
+      })
+    });
+    if (!res.ok) {
+      const body = await res.text();
+      if (body.includes("not found") || body.includes("model")) {
+        console.log(chalk2.yellow(`
+  \u26A0  Chat model "${chatModel}" not found. Pull it: ollama pull ${chatModel}
+`));
+      }
+      return;
+    }
+    const data = await res.json();
+    const raw = data.message.content.trim();
+    let violations = [];
+    try {
+      const parsed = JSON.parse(raw);
+      if (Array.isArray(parsed)) {
+        violations = parsed;
+      } else if (Array.isArray(parsed?.violations)) {
+        violations = parsed.violations;
+      } else if (parsed?.rule) {
+        violations = [parsed];
+      }
+    } catch {
+      violations = [];
+    }
+    if (violations.length === 0) {
+      console.log(chalk2.green(`  \u2713 ${rel}`) + chalk2.dim(" \u2014 no violations"));
+      return;
+    }
+    console.log(
+      chalk2.red.bold(`
+  \u2717 ${violations.length} violation${violations.length > 1 ? "s" : ""} in ${rel}
+`)
+    );
+    violations.forEach((v, i) => {
+      const loc = v.line ? `${v.file ?? rel}:${v.line}` : v.file ?? rel;
+      console.log(chalk2.bold(`  [${i + 1}] ${loc}`));
+      console.log(chalk2.yellow("      Rule:   ") + v.rule);
+      const why = v.reason ?? reasonMap2.get(v.rule);
+      if (why) console.log(chalk2.dim("      Why:    ") + chalk2.dim(why));
+      if (v.issue) console.log(chalk2.red("      Issue:  ") + v.issue);
+      if (v.suggestion) console.log(chalk2.green("      Fix:    ") + v.suggestion);
+      console.log();
+    });
+    console.log(chalk2.dim('  Fix violations or run: memory-core remember "<lesson>"'));
+    console.log();
+  } catch (err) {
+    if (err.cause?.code === "ECONNREFUSED" || err.message?.includes("ECONNREFUSED")) {
+      return;
+    }
+    if (verbose) {
+      console.log(chalk2.yellow(`  \u26A0  Check failed for ${rel}: ${err.message}`));
+    }
+  }
+}
+function startWatch(options = {}) {
+  const cwd = process.cwd();
+  const config2 = loadConfig(cwd);
+  if (!config2) {
+    console.error(chalk2.red("\n  No .memory-core.json found. Run: memory-core init\n"));
+    process.exit(1);
+  }
+  const { rules } = getProfileRules(config2);
+  if (rules.length === 0) {
+    console.log(chalk2.yellow("\n  No architecture rules configured in .memory-core.json \u2014 nothing to watch.\n"));
+    process.exit(0);
+  }
+  const watchPath = options.path ?? cwd;
+  const ollamaUrl = process.env.OLLAMA_URL ?? "http://localhost:11434";
+  const chatModel = process.env.OLLAMA_CHAT_MODEL ?? "llama3.2";
+  console.log(chalk2.cyan("\n  archmind watch \u2014 real-time rule enforcement\n"));
+  console.log(chalk2.dim(`  watching: ${watchPath}`));
+  console.log(chalk2.dim(`  model:    ${chatModel}`));
+  console.log(chalk2.dim(`  rules:    ${rules.length}`));
+  console.log(chalk2.dim("  ctrl+c to stop\n"));
+  const pending = /* @__PURE__ */ new Map();
+  let ollamaWarned = false;
+  const watcher = watch(watchPath, {
+    ignored: [
+      "**/node_modules/**",
+      "**/.git/**",
+      "**/dist/**",
+      "**/build/**",
+      "**/coverage/**",
+      "**/.memory-core*"
+    ],
+    ignoreInitial: true,
+    persistent: true,
+    awaitWriteFinish: { stabilityThreshold: 150, pollInterval: 50 }
+  });
+  const keepAlive = setInterval(() => {
+  }, 1 << 30);
+  const handle = (filePath) => {
+    if (!SOURCE_EXTENSIONS.test(filePath)) return;
+    if (pending.has(filePath)) clearTimeout(pending.get(filePath));
+    const timer = setTimeout(async () => {
+      pending.delete(filePath);
+      console.log(chalk2.dim(`
+  [${(/* @__PURE__ */ new Date()).toLocaleTimeString()}] saved: ${relative(cwd, filePath)}`));
+      try {
+        const ping = await fetch(`${ollamaUrl}/api/tags`, { signal: AbortSignal.timeout(2e3) });
+        if (!ping.ok) throw new Error("not ok");
+        ollamaWarned = false;
+      } catch {
+        if (!ollamaWarned) {
+          console.log(chalk2.yellow(`  \u26A0  Ollama not running at ${ollamaUrl} \u2014 skipping check.`));
+          console.log(chalk2.gray("     Start it: ollama serve\n"));
+          ollamaWarned = true;
+        }
+        return;
+      }
+      await checkFile(filePath, cwd, config2, options.verbose ?? false);
+    }, 300);
+    pending.set(filePath, timer);
+  };
+  watcher.on("add", handle);
+  watcher.on("change", handle);
+  watcher.on("error", (err) => {
+    console.error(chalk2.red(`  watcher error: ${err.message}`));
+  });
+  process.on("SIGINT", () => {
+    console.log(chalk2.dim("\n\n  archmind watch stopped.\n"));
+    clearInterval(keepAlive);
+    watcher.close();
+    process.exit(0);
+  });
+}
+
 // src/cli.ts
 function printBanner(projectName, agentCount) {
   const lines = [
     "",
-    chalk2.cyan("   \u2588\u2588\u2588\u2588\u2588\u2557 \u2588\u2588\u2588\u2588\u2588\u2588\u2557  \u2588\u2588\u2588\u2588\u2588\u2588\u2557\u2588\u2588\u2557  \u2588\u2588\u2557\u2588\u2588\u2588\u2557   \u2588\u2588\u2588\u2557\u2588\u2588\u2557\u2588\u2588\u2588\u2557   \u2588\u2588\u2557\u2588\u2588\u2588\u2588\u2588\u2588\u2557 "),
-    chalk2.cyan("  \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2554\u2550\u2550\u2550\u2550\u255D\u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2557 \u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2557  \u2588\u2588\u2551\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557"),
-    chalk2.cyan("  \u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2588\u2588\u2554\u255D\u2588\u2588\u2551     \u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2554\u2588\u2588\u2588\u2588\u2554\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2554\u2588\u2588\u2557 \u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551"),
-    chalk2.cyan("  \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2551\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2551     \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2551\u2588\u2588\u2551\u255A\u2588\u2588\u2554\u255D\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2551\u255A\u2588\u2588\u2557\u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551"),
-    chalk2.cyan("  \u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551\u255A\u2588\u2588\u2588\u2588\u2588\u2588\u2557\u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2551 \u255A\u2550\u255D \u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2551 \u255A\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2588\u2588\u2554\u255D"),
-    chalk2.cyan("  \u255A\u2550\u255D  \u255A\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u255D \u255A\u2550\u2550\u2550\u2550\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u255D\u255A\u2550\u255D     \u255A\u2550\u255D\u255A\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u2550\u2550\u255D\u255A\u2550\u2550\u2550\u2550\u2550\u255D"),
+    chalk3.cyan("   \u2588\u2588\u2588\u2588\u2588\u2557 \u2588\u2588\u2588\u2588\u2588\u2588\u2557  \u2588\u2588\u2588\u2588\u2588\u2588\u2557\u2588\u2588\u2557  \u2588\u2588\u2557\u2588\u2588\u2588\u2557   \u2588\u2588\u2588\u2557\u2588\u2588\u2557\u2588\u2588\u2588\u2557   \u2588\u2588\u2557\u2588\u2588\u2588\u2588\u2588\u2588\u2557 "),
+    chalk3.cyan("  \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2554\u2550\u2550\u2550\u2550\u255D\u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2557 \u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2557  \u2588\u2588\u2551\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557"),
+    chalk3.cyan("  \u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2588\u2588\u2554\u255D\u2588\u2588\u2551     \u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2554\u2588\u2588\u2588\u2588\u2554\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2554\u2588\u2588\u2557 \u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551"),
+    chalk3.cyan("  \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2551\u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2557\u2588\u2588\u2551     \u2588\u2588\u2554\u2550\u2550\u2588\u2588\u2551\u2588\u2588\u2551\u255A\u2588\u2588\u2554\u255D\u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2551\u255A\u2588\u2588\u2557\u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551"),
+    chalk3.cyan("  \u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2551  \u2588\u2588\u2551\u255A\u2588\u2588\u2588\u2588\u2588\u2588\u2557\u2588\u2588\u2551  \u2588\u2588\u2551\u2588\u2588\u2551 \u255A\u2550\u255D \u2588\u2588\u2551\u2588\u2588\u2551\u2588\u2588\u2551 \u255A\u2588\u2588\u2588\u2588\u2551\u2588\u2588\u2588\u2588\u2588\u2588\u2554\u255D"),
+    chalk3.cyan("  \u255A\u2550\u255D  \u255A\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u255D \u255A\u2550\u2550\u2550\u2550\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u255D\u255A\u2550\u255D     \u255A\u2550\u255D\u255A\u2550\u255D\u255A\u2550\u255D  \u255A\u2550\u2550\u2550\u255D\u255A\u2550\u2550\u2550\u2550\u2550\u255D"),
     "",
-    chalk2.dim("  \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500 ") + chalk2.bold.white("C O R E") + chalk2.dim(" \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500"),
+    chalk3.dim("  \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500 ") + chalk3.bold.white("C O R E") + chalk3.dim(" \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500"),
     "",
-    chalk2.green(`  \u2713 Project   `) + chalk2.bold(projectName),
-    chalk2.green(`  \u2713 Agents    `) + chalk2.bold(`${agentCount} AI agents configured`),
-    chalk2.green(`  \u2713 Memory    `) + chalk2.bold("PostgreSQL + pgvector ready"),
+    chalk3.green(`  \u2713 Project   `) + chalk3.bold(projectName),
+    chalk3.green(`  \u2713 Agents    `) + chalk3.bold(`${agentCount} AI agents configured`),
+    chalk3.green(`  \u2713 Memory    `) + chalk3.bold("PostgreSQL + pgvector ready"),
     "",
-    chalk2.dim("  \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500"),
+    chalk3.dim("  \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500"),
     "",
-    chalk2.bold("  Every AI agent in this project now follows your rules."),
+    chalk3.bold("  Every AI agent in this project now follows your rules."),
     "",
-    chalk2.gray("  Next steps:"),
-    chalk2.gray('    memory-core remember "Your architectural decision"'),
-    chalk2.gray('    memory-core search "query"'),
-    chalk2.gray("    memory-core sync"),
+    chalk3.gray("  Next steps:"),
+    chalk3.gray('    memory-core remember "Your architectural decision"'),
+    chalk3.gray('    memory-core search "query"'),
+    chalk3.gray("    memory-core sync"),
     ""
   ];
   lines.forEach((l) => console.log(l));
 }
 var CONFIG_FILE = ".memory-core.json";
 function readProjectConfig() {
-  const path = join5(process.cwd(), CONFIG_FILE);
-  if (!existsSync5(path)) return null;
+  const path = join6(process.cwd(), CONFIG_FILE);
+  if (!existsSync6(path)) return null;
   try {
-    return JSON.parse(readFileSync4(path, "utf-8"));
+    return JSON.parse(readFileSync5(path, "utf-8"));
   } catch {
     return null;
   }
 }
 function writeProjectConfig(config2) {
-  writeFileSync3(join5(process.cwd(), CONFIG_FILE), JSON.stringify(config2, null, 2));
+  writeFileSync3(join6(process.cwd(), CONFIG_FILE), JSON.stringify(config2, null, 2));
 }
 var program = new Command();
 program.name("memory-core").description("Universal AI memory core \u2014 generate AI context files for all coding agents").version("0.1.0");
 program.command("init").description("Initialize memory-core in the current project").action(async () => {
-  console.log(chalk2.bold.cyan("\n  memory-core init\n"));
+  console.log(chalk3.bold.cyan("\n  memory-core init\n"));
   const detected = detectProject();
-  const envPath = join5(process.cwd(), ".memory-core.env");
-  const hasEnv = existsSync5(envPath) || existsSync5(join5(process.cwd(), ".env")) || !!process.env.DATABASE_URL;
+  const envPath = join6(process.cwd(), ".memory-core.env");
+  const hasEnv = existsSync6(envPath) || existsSync6(join6(process.cwd(), ".env")) || !!process.env.DATABASE_URL;
   if (!hasEnv) {
-    console.log(chalk2.dim("  No .memory-core.env found \u2014 let's set up your database connection.\n"));
+    console.log(chalk3.dim("  No .memory-core.env found \u2014 let's set up your database connection.\n"));
     const dbUser = process.env.USER ?? process.env.USERNAME ?? "postgres";
     const dbUrl = await input({
       message: "PostgreSQL connection URL?",
@@ -853,17 +1079,17 @@ program.command("init").description("Initialize memory-core in the current proje
     process.env.OLLAMA_URL = ollamaUrl;
     process.env.OLLAMA_MODEL = "nomic-embed-text";
     process.env.OLLAMA_CHAT_MODEL = "llama3.2";
-    const gitignorePath = join5(process.cwd(), ".gitignore");
-    if (existsSync5(gitignorePath)) {
-      const gi = readFileSync4(gitignorePath, "utf-8");
+    const gitignorePath = join6(process.cwd(), ".gitignore");
+    if (existsSync6(gitignorePath)) {
+      const gi = readFileSync5(gitignorePath, "utf-8");
       if (!gi.includes(".memory-core.env")) {
         appendFileSync(gitignorePath, "\n.memory-core.env\n");
       }
     } else {
       writeFileSync3(gitignorePath, ".memory-core.env\n");
     }
-    console.log(chalk2.green("\n  \u2713 .memory-core.env created"));
-    console.log(chalk2.gray("    Added to .gitignore \u2014 your DB credentials stay local.\n"));
+    console.log(chalk3.green("\n  \u2713 .memory-core.env created"));
+    console.log(chalk3.gray("    Added to .gitignore \u2014 your DB credentials stay local.\n"));
   }
   const projectName = await input({
     message: "Project name?",
@@ -944,7 +1170,7 @@ program.command("init").description("Initialize memory-core in the current proje
   if (installCaveman) {
     const spinner2 = ora("Installing caveman token saver\u2026").start();
     try {
-      execSync2(
+      execSync3(
         "curl -fsSL https://raw.githubusercontent.com/JuliusBrussee/caveman/main/install.sh | bash",
         { stdio: "pipe", cwd: process.cwd() }
       );
@@ -966,7 +1192,7 @@ program.command("init").description("Initialize memory-core in the current proje
 program.command("sync").description("Re-pull memories and regenerate all AI agent files").action(async () => {
   const config2 = readProjectConfig();
   if (!config2) {
-    console.error(chalk2.red("No .memory-core.json found. Run: memory-core init"));
+    console.error(chalk3.red("No .memory-core.json found. Run: memory-core init"));
     process.exit(1);
   }
   const spinner = ora("Syncing memories\u2026").start();
@@ -998,7 +1224,7 @@ program.command("remember <text>").description("Save a new memory to the central
   let reason = opts.reason;
   if (!reason) {
     reason = await input({
-      message: chalk2.dim("Why does this rule exist? (optional \u2014 helps agents debug violations)"),
+      message: chalk3.dim("Why does this rule exist? (optional \u2014 helps agents debug violations)"),
       default: ""
     });
   }
@@ -1015,9 +1241,9 @@ program.command("remember <text>").description("Save a new memory to the central
       tags: opts.tags ? opts.tags.split(",").map((t) => t.trim()) : [],
       embedding
     });
-    const reasonLine = reason ? chalk2.gray(`
+    const reasonLine = reason ? chalk3.gray(`
   Why: ${reason}`) : "";
-    spinner.succeed(chalk2.green(`Memory saved: "${text}"`) + reasonLine);
+    spinner.succeed(chalk3.green(`Memory saved: "${text}"`) + reasonLine);
   } catch (err) {
     spinner.fail(`Failed: ${err.message}`);
     process.exit(1);
@@ -1035,16 +1261,16 @@ program.command("search <query>").description("Search memories using semantic si
     );
     spinner.stop();
     if (results.length === 0) {
-      console.log(chalk2.yellow("No memories found."));
+      console.log(chalk3.yellow("No memories found."));
     } else {
-      console.log(chalk2.bold(`
+      console.log(chalk3.bold(`
   ${results.length} results for "${query}"
 `));
       results.forEach((m, i) => {
-        const sim = m.similarity ? chalk2.gray(` (${(m.similarity * 100).toFixed(0)}% match)`) : "";
-        console.log(chalk2.cyan(`  ${i + 1}. [${m.type}] ${m.title ?? ""}`));
-        console.log(chalk2.white(`     ${m.content}`) + sim);
-        if (m.tags?.length) console.log(chalk2.gray(`     tags: ${m.tags.join(", ")}`));
+        const sim = m.similarity ? chalk3.gray(` (${(m.similarity * 100).toFixed(0)}% match)`) : "";
+        console.log(chalk3.cyan(`  ${i + 1}. [${m.type}] ${m.title ?? ""}`));
+        console.log(chalk3.white(`     ${m.content}`) + sim);
+        if (m.tags?.length) console.log(chalk3.gray(`     tags: ${m.tags.join(", ")}`));
         console.log();
       });
     }
@@ -1057,7 +1283,7 @@ program.command("search <query>").description("Search memories using semantic si
 program.command("seed").description("Load all predefined memories into the database").option("--arch <architecture>", "Only seed a specific architecture (e.g. clean-architecture)").option("--force", "Re-seed even if memories already exist", false).action(async (opts) => {
   await runMigrations();
   const filtered = opts.arch ? seeds.filter((s) => s.architecture === opts.arch || s.architecture === "global") : seeds;
-  console.log(chalk2.bold.cyan(`
+  console.log(chalk3.bold.cyan(`
   Seeding ${filtered.length} memories\u2026
 `));
   let saved = 0;
@@ -1076,14 +1302,14 @@ program.command("seed").description("Load all predefined memories into the datab
         tags: seed.tags,
         embedding
       });
-      spinner.succeed(chalk2.gray(`[${seed.architecture}] ${seed.title}`));
+      spinner.succeed(chalk3.gray(`[${seed.architecture}] ${seed.title}`));
       saved++;
     } catch (err) {
       spinner.warn(`Skipped \u2014 ${err.message}`);
       skipped++;
     }
   }
-  console.log(chalk2.bold.green(`
+  console.log(chalk3.bold.green(`
   Done. ${saved} memories seeded, ${skipped} skipped.
 `));
   await closePool();
@@ -1092,21 +1318,21 @@ program.command("global").description("Sync your memory into every AI agent glob
   const home = homedir();
   const GLOBAL_TARGETS = [
     // Claude Code
-    { label: "Claude Code", path: join5(home, ".claude/CLAUDE.md"), type: "md" },
+    { label: "Claude Code", path: join6(home, ".claude/CLAUDE.md"), type: "md" },
     // GitHub Copilot (VS Code)
-    { label: "Copilot", path: join5(home, "Library/Application Support/Code/User/settings.json"), type: "vscode-copilot" },
+    { label: "Copilot", path: join6(home, "Library/Application Support/Code/User/settings.json"), type: "vscode-copilot" },
     // Cursor global rules
-    { label: "Cursor", path: join5(home, ".cursor/rules/memory-core.mdc"), type: "md" },
+    { label: "Cursor", path: join6(home, ".cursor/rules/memory-core.mdc"), type: "md" },
     // Cline (VS Code)
-    { label: "Cline", path: join5(home, "Library/Application Support/Code/User/settings.json"), type: "vscode-cline" },
+    { label: "Cline", path: join6(home, "Library/Application Support/Code/User/settings.json"), type: "vscode-cline" },
     // Continue.dev global config
-    { label: "Continue.dev", path: join5(home, ".continue/config.json"), type: "continue" },
+    { label: "Continue.dev", path: join6(home, ".continue/config.json"), type: "continue" },
     // Aider global config
-    { label: "Aider", path: join5(home, ".aider.conf.yml"), type: "aider" },
+    { label: "Aider", path: join6(home, ".aider.conf.yml"), type: "aider" },
     // Zed global settings
-    { label: "Zed AI", path: join5(home, ".config/zed/settings.json"), type: "zed" },
+    { label: "Zed AI", path: join6(home, ".config/zed/settings.json"), type: "zed" },
     // Windsurf global rules
-    { label: "Windsurf", path: join5(home, ".windsurf/rules/memory-core.md"), type: "md" }
+    { label: "Windsurf", path: join6(home, ".windsurf/rules/memory-core.md"), type: "md" }
   ];
   const spinner = ora("Fetching global memories\u2026").start();
   let memories = [];
@@ -1134,9 +1360,9 @@ ${rulesText}
     writeFileSync3(filePath, content, "utf-8");
   };
   const readJson = (filePath) => {
-    if (!existsSync5(filePath)) return {};
+    if (!existsSync6(filePath)) return {};
     try {
-      return JSON.parse(readFileSync4(filePath, "utf-8"));
+      return JSON.parse(readFileSync5(filePath, "utf-8"));
     } catch {
       return {};
     }
@@ -1184,14 +1410,14 @@ read:
       skipped.push(target.label);
     }
   }
-  spinner.succeed(chalk2.green(`Synced ${memories.length} memories \u2192 ${written.length} agents`));
-  console.log(chalk2.green("\n  Updated:"));
-  written.forEach((l) => console.log(chalk2.gray(`    \u2713 ${l}`)));
+  spinner.succeed(chalk3.green(`Synced ${memories.length} memories \u2192 ${written.length} agents`));
+  console.log(chalk3.green("\n  Updated:"));
+  written.forEach((l) => console.log(chalk3.gray(`    \u2713 ${l}`)));
   if (skipped.length) {
-    console.log(chalk2.yellow("\n  Skipped (not installed):"));
-    skipped.forEach((l) => console.log(chalk2.gray(`    \u2717 ${l}`)));
+    console.log(chalk3.yellow("\n  Skipped (not installed):"));
+    skipped.forEach((l) => console.log(chalk3.gray(`    \u2717 ${l}`)));
   }
-  console.log(chalk2.bold("\n  Every AI agent now follows your memory globally.\n"));
+  console.log(chalk3.bold("\n  Every AI agent now follows your memory globally.\n"));
   await closePool();
 });
 var hook = program.command("hook").description("Manage the pre-commit rule enforcement hook");
@@ -1204,4 +1430,7 @@ hook.command("uninstall").description("Remove the pre-commit hook").action(() =>
 program.command("check").description("Check staged changes against architecture rules (used by pre-commit hook)").option("--staged", "Check git staged diff (default behaviour)").option("--verbose", "Show model and diff details").action(async (opts) => {
   await checkStaged({ verbose: opts.verbose ?? false });
 });
+program.command("watch").description("Watch source files and check violations in real-time on every save").option("--path <dir>", "Directory to watch (default: current directory)").option("--verbose", "Show diff size and model details per file").action((opts) => {
+  startWatch({ path: opts.path, verbose: opts.verbose });
+});
 program.parseAsync(process.argv);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@shahmilsaari/memory-core",
-  "version": "0.1.2",
+  "version": "0.1.4",
   "description": "Universal AI memory core — generate AI context files from architecture profiles with RAG support",
   "type": "module",
   "bin": {
@@ -19,6 +19,7 @@
   "dependencies": {
     "@inquirer/prompts": "^5.0.0",
     "chalk": "^5.3.0",
+    "chokidar": "^5.0.0",
     "commander": "^12.0.0",
     "dotenv": "^16.4.0",
     "handlebars": "^4.7.8",