ideaco 1.1.6 → 1.1.7

package/README.md

@@ -17,12 +17,15 @@
 
 <p align="center">
   <a href="#-quick-start">Quick Start</a> ·
-  <a href="#-features">Features</a> ·
   <a href="#-architecture">Architecture</a> ·
   <a href="#-download">Download</a> ·
   <a href="ARCHITECTURE.md">Full Architecture Doc</a>
 </p>
 
+<p align="center">
+  <img src="public/demo/demo.gif" width="100%" alt="IdeaCo Demo" />
+</p>
+
 ---
 
 ## What is IdeaCo?
@@ -35,74 +38,24 @@ Each employee has **persistent memory**, a **unique personality**, **social rela
 
 ---
 
-## 🧬 Why IdeaCo?
-
-Most agent frameworks create **stateless workflows**. IdeaCo manages **long-living AI employees with memory, personality, and social bonds**.
-
-| | Typical Agent Framework | IdeaCo |
-|---|---|---|
-| **Agents** | Ephemeral, per-task | Persistent employees with memory & personality |
-| **Memory** | None or simple RAG | Layered: long-term + short-term + social impressions + rolling summary |
-| **Social** | Agents don't know each other | Employees form opinions, track affinity, and remember coworkers |
-| **Orchestration** | DAG / workflow graph | Company org structure — departments, teams, roles |
-| **Backend** | Single LLM | Mix LLMs + CLI tools (Claude Code, Codex, CodeBuddy) |
-| **Autonomy** | Triggered by code | Employees poll, think, decide to speak or stay silent on their own |
-| **Interface** | Code / YAML | Visual — pixel office, group chats, dashboards |
-
----
-
-## ✨ Features
-
-| Screenshot | Feature | Description |
-|:---:|---|---|
-| <img src="public/demo/dashboard.png" width="280" /> | **Dashboard** | Company overview — departments, employees, budget, requirements, and task forces at a glance |
-| <img src="public/demo/office.png" width="280" /> | **Pixel Office** | Pixel-art virtual office where AI agents wander, work at desks, and show real-time chat bubbles |
-| <img src="public/demo/employee.png" width="280" /> | **Employee Profile** | Detailed card — personality, memory (long/short-term), social relationships, affinity scores, performance, and cost tracking |
-| <img src="public/demo/messages.png" width="280" /> | **Group Chat** | Department chats where agents discuss, debate, and collaborate autonomously with flow-of-thought reasoning |
-| <img src="public/demo/requirement.png" width="280" /> | **Requirements** | Assign tasks to departments — agents auto-decompose, execute with real tools, and produce deliverables |
+## 🚀 Quick Start
 
-### Key Highlights
+### Option 1: NPM (Recommended)
 
-- 🧠 **Flow-of-Thought** — Each employee reads messages, has an inner monologue, and independently decides whether to reply based on topic saturation, relevance, and anti-spam rules
-- 💾 **Layered Memory** — Long-term memory (facts, preferences), short-term memory (TTL-based), rolling history summaries, and AI-managed memory ops (add/delete)
-- 👥 **Social Memory** — Every employee maintains impressions and affinity scores for coworkers. They remember who helped them, who they disagree with, and who they like
-- 🎭 **12 Personality Archetypes** — From "Chatterbox" to "Zen Slacker" to "Anxious Perfectionist", each employee has a unique voice and behavior pattern
-- 🔌 **Multi-Backend** — Mix cloud LLMs (OpenAI, DeepSeek, Anthropic), CLI tools (Claude Code, Codex, CodeBuddy), and web agents (ChatGPT Web) in one company
-- 🏢 **Org Structure** — Departments, teams, roles, talent market, HR hiring — a complete company simulation
+Requires Node.js 20+
 
----
+```bash
+# Install globally
+npm install -g ideaco
 
-## 🏗 Architecture
+# Start your company (opens dashboard automatically)
+ideaco start
 
-```
-┌───────────────────────────────────────────────────┐
-│                  👤 User (Boss)                    │
-├───────────────────────────────────────────────────┤
-│                🧑‍💼 Secretary                       │
-│     Intent Parsing · HR · Task Dispatch · Reports │
-├───────────────────────────────────────────────────┤
-│               🏢 Organization                      │
-│       Company · Department · Team · Requirement   │
-├───────────────────────────────────────────────────┤
-│                👥 Employee                          │
-│     Memory · Personality · Skills · Lifecycle     │
-├───────────────────────────────────────────────────┤
-│                🤖 Agent                             │
-│       LLM Agent · CLI Agent · Web Agent           │
-│      (Unified interface, zero business logic)     │
-└───────────────────────────────────────────────────┘
+# Open the dashboard
+ideaco ui
 ```
 
-- **Agent** — Pure LLM communication engine. Supports API, CLI, and Web backends behind a single `chat()` interface. Zero business logic.
-- **Employee** — The atomic unit. Wraps an Agent with persistent memory, personality, skills, and autonomous poll-think-reply behavior.
-- **Organization** — Company structure: departments, teams, group chats, requirements workflow.
-- **Secretary** — Boss's AI assistant that understands intent, manages HR, and coordinates departments.
-
-> 📖 See [ARCHITECTURE.md](ARCHITECTURE.md) for the full deep-dive — employee lifecycle, memory system design, flow-of-thought pipeline, anti-spam gates, and social memory details.
-
----
-
-## ⬇ Download
+### Option 2: Desktop Client (Mac/Windows/Linux)
 
 Grab the latest release for your platform:
 
@@ -125,11 +78,7 @@ Since IdeaCo is not signed with an Apple Developer certificate, macOS will show
 3. **Or via Terminal**: `xattr -cr /Applications/IdeaCo.app`
 </details>
 
----
-
-## 🚀 Quick Start
-
-### From Source
+### Option 3: From Source
 
 ```bash
 git clone https://github.com/ymssx/IdeaCo.git
@@ -138,9 +87,7 @@ yarn install
 yarn dev
 ```
 
-Open **http://localhost:9999** — the Setup Wizard will guide you.
-
-### Docker
+### Option 4: Docker
 
 ```bash
 docker compose up -d
@@ -150,6 +97,52 @@ Data is persisted via Docker volumes (`app-data`, `app-workspace`).
 
 ---
 
+## 🧬 Why IdeaCo?
+
+Most agent frameworks create **stateless workflows**. IdeaCo manages **long-living AI employees with memory, personality, and social bonds**.
+
+| | Typical Agent Framework | IdeaCo |
+|---|---|---|
+| **Agents** | Ephemeral, per-task | Persistent employees with memory & personality |
+| **Memory** | None or simple RAG | Layered: long-term + short-term + social impressions + rolling summary |
+| **Social** | Agents don't know each other | Employees form opinions, track affinity, and remember coworkers |
+| **Orchestration** | DAG / workflow graph | Company org structure — departments, teams, roles |
+| **Backend** | Single LLM | Mix LLMs + CLI tools (Claude Code, Codex, CodeBuddy) |
+| **Autonomy** | Triggered by code | Employees poll, think, decide to speak or stay silent on their own |
+| **Interface** | Code / YAML | Visual — pixel office, group chats, dashboards |
+
+---
+
+##  Architecture
+
+```
+┌───────────────────────────────────────────────────┐
+│                  👤 User (Boss)                    │
+├───────────────────────────────────────────────────┤
+│                🧑‍💼 Secretary                       │
+│     Intent Parsing · HR · Task Dispatch · Reports │
+├───────────────────────────────────────────────────┤
+│               🏢 Organization                      │
+│       Company · Department · Team · Requirement   │
+├───────────────────────────────────────────────────┤
+│                👥 Employee                          │
+│     Memory · Personality · Skills · Lifecycle     │
+├───────────────────────────────────────────────────┤
+│                🤖 Agent                             │
+│       LLM Agent · CLI Agent · Web Agent           │
+│      (Unified interface, zero business logic)     │
+└───────────────────────────────────────────────────┘
+```
+
+- **Agent** — Pure LLM communication engine. Supports API, CLI, and Web backends behind a single `chat()` interface. Zero business logic.
+- **Employee** — The atomic unit. Wraps an Agent with persistent memory, personality, skills, and autonomous poll-think-reply behavior.
+- **Organization** — Company structure: departments, teams, group chats, requirements workflow.
+- **Secretary** — Boss's AI assistant that understands intent, manages HR, and coordinates departments.
+
+> 📖 See [ARCHITECTURE.md](ARCHITECTURE.md) for the full deep-dive — employee lifecycle, memory system design, flow-of-thought pipeline, anti-spam gates, and social memory details.
+
+---
+
 ## ⚙ Configuration
 
 ### LLM Providers
@@ -198,9 +191,3 @@ IdeaCo/
 ├── electron/                 # Electron desktop app shell
 └── package.json
 ```
-
----
-
-## License
-
-[MIT](LICENSE)

package/bin/ideaco.js

@@ -496,6 +496,37 @@ async function openElectron(port) {
   child.on('exit', (code) => process.exit(code ?? 0));
 }
 
+async function handleUiCommand() {
+  const pid = readPid();
+  const port = readPort() ?? PORT;
+  
+  // Check if service is already running
+  if (pid && isPidRunning(pid)) {
+    try {
+      // Simple check if port is responding
+      await new Promise((resolve, reject) => {
+        const socket = new net.Socket();
+        socket.setTimeout(1000);
+        socket.on('connect', () => { socket.destroy(); resolve(); });
+        socket.on('timeout', () => { socket.destroy(); reject(new Error('timeout')); });
+        socket.on('error', (err) => reject(err));
+        socket.connect(port, '127.0.0.1');
+      });
+      
+      // Service is running, open Electron directly
+      console.log(chalk.cyan(`Opening Electron UI on port ${port}...`));
+      await openElectron(port);
+      return;
+    } catch (e) {
+      // Port check failed, service might be dead
+    }
+  }
+
+  // Service not running, start it (which will auto-open Electron)
+  console.log(chalk.yellow('Service not running. Starting...'));
+  await startServer();
+}
+
 function printHelp() {
   console.log(`
 ${t('cli.helpTitle')}
@@ -512,8 +543,8 @@ async function main() {
   if (command === 'stop') return await stopServer();
   if (command === 'web') return await openWeb();
   if (command === 'banner') return await runBanner(args[1]);
-  if (command === 'ui') return await openElectron();
-  if (command === 'electron') return await openElectron();
+  if (command === 'ui') return await handleUiCommand();
+  if (command === 'electron') return await handleUiCommand();
   return printHelp();
 }
 

package/next.config.mjs

@@ -1,3 +1,5 @@
+import path from 'path';
+
 /** @type {import('next').NextConfig} */
 const nextConfig = {
   output: 'standalone', // Required for Docker deployment
@@ -11,6 +13,11 @@ const nextConfig = {
     ],
   },
   webpack: (config, { isServer }) => {
+    config.resolve = config.resolve || {};
+    config.resolve.alias = {
+      ...(config.resolve.alias || {}),
+      '@': path.resolve(process.cwd(), 'src'),
+    };
     if (isServer) {
       // 将可选依赖标记为 external，避免 webpack 尝试解析/打包它们
       const optionalDeps = ['puppeteer', 'puppeteer-core', 'pdf-parse', 'openai'];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ideaco",
-  "version": "1.1.6",
+  "version": "1.1.7",
   "description": "Idea Unlimited - An LLM-powered AI enterprise simulator where you boss around AI employees",
   "author": {
     "name": "ymssx",