ideaco 1.1.5

package/.dockerignore

@@ -0,0 +1,33 @@
+# Docker-related
+.dockerignore
+*.log
+Dockerfile
+docker-compose.yml
+
+# Dependencies (will be installed in build)
+node_modules/
+
+# Build output (will be rebuilt)
+.next/
+
+# Environment secrets
+.env
+.env.local
+.env.*.local
+
+# Development data
+data/
+workspace/
+
+# Git
+.git/
+.gitmodules
+
+# Reference submodule (not needed in production)
+vendor/
+
+# Documentation
+*.md
+LICENSE
+DISTILLATION-MANIFESTO.md
+THIRD-PARTY-NOTICES.md

package/.nvmrc

@@ -0,0 +1 @@
+20

package/ARCHITECTURE.md

@@ -0,0 +1,394 @@
+# IdeaCo Architecture
+
+> Internal architecture documentation for IdeaCo — the AI Employee Management System.
+
+---
+
+## System Overview
+
+IdeaCo is a layered AI enterprise simulator where every component maps to a real-world company metaphor. The system is organized into five clear layers:
+
+```
+┌─────────────────────────────────────────────────────┐
+│                   👤 User (Boss)                     │
+│              Chat interface / Mailbox                │
+├─────────────────────────────────────────────────────┤
+│                 🧑‍💼 Secretary                        │
+│      Intent parsing · HR coordination · Reporting   │
+├─────────────────────────────────────────────────────┤
+│              🏢 Organization Layer                   │
+│     Company · Department · Team · Requirement       │
+├─────────────────────────────────────────────────────┤
+│               👥 Employee Layer                      │
+│   Memory · Personality · Skills · Prompt · Lifecycle│
+├─────────────────────────────────────────────────────┤
+│               🤖 Agent Layer                         │
+│        LLM Agent · CLI Agent · Web Agent            │
+│       (Unified interface, zero business logic)      │
+└─────────────────────────────────────────────────────┘
+```
+
+### Layer Responsibilities
+
+| Layer | Module | Responsibility |
+|-------|--------|----------------|
+| **User** | Frontend / Mailbox | Boss gives orders, views progress, and manages the company |
+| **Secretary** | `secretary.js` | Understands boss intent, coordinates HR, dispatches tasks, reports progress |
+| **Organization** | `company.js` / `department.js` / `team.js` / `requirement.js` | Company structure, department lifecycle, requirement workflow, group chat containers |
+| **Employee** | `base-employee.js` / `lifecycle.js` / `memory/` | The atomic management unit — wraps an Agent with memory, personality, skills, and autonomous behavior |
+| **Agent** | `base-agent.js` → LLMAgent / CLIAgent / WebAgent | Pure LLM communication engine. Zero business logic. Uniform `chat()` interface |
+
+---
+
+## Agent Layer
+
+The Agent layer is a **pure communication engine** — it only handles LLM input/output and is completely decoupled from business logic. All three agent types expose the same abstract interface:
+
+```
+BaseAgent (abstract)
+  ├── LLMAgent     — OpenAI-compatible API (API key auth)
+  ├── CLIAgent     — Local CLI tools (Claude Code, Codex, CodeBuddy)
+└── WebAgent     — Browser DOM automation (ChatGPT Web, Claude Web)
+```
+
+**Unified Interface:**
+- `chat(messages, options)` — Send a conversation request
+- `chatWithTools(messages, toolExecutor, options)` — Conversation with tool calling
+- `isAvailable()` / `canChat()` — Availability check
+- `switchProvider(newProvider)` — Runtime provider switching
+- `serialize()` / `deserialize()` — State persistence
+
+**Internal differentiation (not exposed):**
+- **LLMAgent**: Native function calling via OpenAI SDK
+- **CLIAgent**: Subprocess execution, stdout parsing, fallback to LLM
+- **WebAgent**: Hidden BrowserWindow DOM scripting, per-employee session isolation, tool-call simulation via prompt
+
+---
+
+## Employee Layer
+
+An Employee is an Agent wrapped with **identity, memory, personality, and autonomous behavior**. It is the atomic management unit of the system.
+
+### Employee Lifecycle
+
+```
+┌──────────┐    ┌───────────┐    ┌─────────┐    ┌──────────────┐
+│ 1. Init  │───▶│2. Onboard │───▶│3. WakeUp│───▶│ 4. Poll Loop │◀─┐
+└──────────┘    └───────────┘    └─────────┘    └──────┬───────┘  │
+                                                       │          │
+                                                       ▼          │
+                                              ┌────────────────┐  │
+                                              │ 5. Read Group  │  │
+                                              │    Messages    │  │
+                                              └───────┬────────┘  │
+                                                      │           │
+                                                      ▼           │
+                                              ┌────────────────┐  │
+                                              │ 6. Flow-of-    │  │
+                                              │   Thought      │  │
+                                              └───────┬────────┘  │
+                                                      │           │
+                                                ┌─────┴─────┐    │
+                                                ▼           ▼    │
+                                          ┌─────────┐ ┌────────┐ │
+                                          │  Reply  │ │ Silent │ │
+                                          └────┬────┘ └────┬───┘ │
+                                               │          │      │
+                                               ▼          │      │
+                                          ┌──────────┐    │      │
+                                          │ Trigger  │    │      │
+                                          │  Others  │────┘──────┘
+                                          └──────────┘
+```
+
+#### Phase 1 — Initialization (Constructor)
+
+- **Identity**: Name, gender, age, role, skill set
+- **Agent Binding**: Attach an Agent instance (LLM / CLI / Web)
+- **Personality Template**: Randomly assigned from 12 archetype pool (e.g., "Chatterbox", "Zen Slacker", "Anxious Perfectionist")
+- **Memory System**: Initialize Memory instance (long-term + short-term + history summary + social impressions)
+- **Toolbox**: Initialize AgentToolKit (shell, file I/O, search, etc.)
+
+#### Phase 2 — Onboarding
+
+The employee uses its own LLM capability to generate a self-introduction:
+
+1. Build a prompt containing its persona settings (name, role, personality archetype)
+2. Call `chat()` to have the LLM generate:
+   - **Self-introduction** — a paragraph about themselves
+   - **Signature** — a personal tagline
+   - **Personality description** — behavioral traits
+3. Persist generated content to employee state
+4. Send a greeting message to the Boss (`sendMailToBoss`)
+
+#### Phase 3 — Wake Up
+
+Triggered on app startup or session refresh. Injects full identity into the Agent's conversation context:
+
+- **System Message**: Personal info, behavioral norms, skill list
+- **Long-term Memory**: Top 20 entries by importance
+- **Short-term Memory**: Top 10 active (non-expired) entries
+- Calls `agent.chat()` to warm up the session — the Agent "remembers who it is"
+
+#### Phase 4 — Poll Cycle
+
+Each employee runs an independent random-interval timer:
+
+- **Interval**: 10 seconds ~ 5 minutes (simulates human-like irregular checking behavior)
+- **Each Cycle**: Scan all joined groups (work groups + lounge groups), check for unread messages
+- **Lounge Groups**: After idle threshold, 15% chance of initiating a topic spontaneously
+
+#### Phase 5 — Read Group Messages
+
+When unread messages are detected:
+
+1. Separate **read messages** (old context) from **unread messages** (new, requiring response)
+2. Wait 5–20s random jitter (simulates "thinking/typing" time)
+3. Any messages arriving during jitter are absorbed into the unread batch
+4. Enter flow-of-thought phase
+
+#### Phase 6 — Flow-of-Thought (Agent Think)
+
+Constructs a full prompt context and calls the LLM for decision-making.
+
+**Input (injected into LLM):**
+
+| # | Input | Description |
+|---|-------|-------------|
+| 1 | Personal Info | Name, role, personality, speaking style |
+| 2 | Settings & Abilities | Behavioral norms, role definition, skills |
+| 3 | Scene Context | Group name, members, department mission, requirement description |
+| 4 | Chat History | Last 20 messages (read/unread markers) |
+| 5 | Memory | Rolling history summary + Long-term (top 15) + Short-term (top 10) |
+| 6 | Social Memory | Relationship impressions of message senders (name, impression, affinity) |
+| 7 | Dedup Hint | If colleagues already replied with similar content → "don't repeat others" |
+| 8 | Perspective Seed | Random thinking angle ("pragmatic", "innovative", "risk-focused") to prevent opinion convergence |
+
+**LLM Structured Output (JSON):**
+
+```json
+{
+  "innerThoughts": "Internal monologue with personality and emotion",
+  "topicSaturation": 5,
+  "shouldSpeak": true,
+  "reason": "Why I decided to speak/stay silent",
+  "messages": [{ "content": "The actual reply" }],
+  "memorySummary": "Compressed summary of old messages",
+  "memoryOps": [
+    { "op": "add", "type": "long_term", "content": "...", "category": "fact", "importance": 8 },
+    { "op": "add", "type": "short_term", "content": "...", "ttl": 3600 },
+    { "op": "delete", "id": "mem_xxx" }
+  ],
+  "relationshipOps": [
+    { "employeeId": "emp_123", "name": "Alice", "impression": "Strong coder, reliable", "affinity": 75 }
+  ]
+}
+```
+
+### Reply Decision Gates
+
+After the LLM's flow-of-thought, the reply passes through multiple filter gates:
+
+```
+LLM returns shouldSpeak=true
+        │
+        ▼
+┌─────────────────────┐     ┌──────────┐
+│ Topic Saturation ≥7?│──▶  │ 🔇 Silent│
+└─────────┬───────────┘     └──────────┘
+          │ No
+          ▼
+┌─────────────────────┐     ┌──────────┐
+│ Anti-spam: Too many │──▶  │ 🔇 Silent│
+│ messages in window? │     └──────────┘
+└─────────┬───────────┘
+          │ No
+          ▼
+┌─────────────────────┐     ┌──────────┐
+│ Cooldown active?    │──▶  │ 🔇 Silent│
+└─────────┬───────────┘     └──────────┘
+          │ No
+          ▼
+     ┌──────────┐
+     │ ✅ Speak │
+     └──────────┘
+```
+
+| Gate | Mechanism | Parameters |
+|------|-----------|------------|
+| **Topic Saturation** | LLM self-rates 1–10; ≥ 7 forces silence | threshold = 7 |
+| **Anti-spam (Work)** | Max 2 messages per 5-minute window | window=5min, max=2 |
+| **Anti-spam (Lounge)** | Max 4 messages per 10-minute window | window=10min, max=4 |
+| **Cooldown (Work)** | 60s silence after speaking | cooldown=60s |
+| **Cooldown (Lounge)** | 30s silence after speaking | cooldown=30s |
+
+### Chain Reaction
+
+After an employee sends a message:
+
+1. Message is written to the group chat log
+2. Send timestamp is recorded (for anti-spam)
+3. **Nudge all other group members** with random delay → triggers their `_processGroupMessages`, creating an autonomous conversation loop
+
+---
+
+## Memory System
+
+### Memory Architecture
+
+```
+Memory
+  ├── 📜 historySummary    — Rolling AI-compressed summary of old messages (per-group)
+  ├── 💾 longTerm          — Permanent storage, sorted by importance (1–10), cap: 200
+  ├── ⚡ shortTerm          — TTL-based auto-expiry, default 24h, cap: 20
+  └── 👥 relationships     — Social impression table: Map<employeeId, {name, impression, affinity}>
+```
+
+### Per-Conversation Memory Input
+
+| # | Input | Description |
+|---|-------|-------------|
+| 1 | Personal Info | Name, role, age, gender, personality, signature |
+| 2 | Settings & Abilities | System prompt, skill list, behavioral norms |
+| 3 | Scene Context | Group name, members, department mission, requirement |
+| 4 | Chat History | Last 20 messages (read/unread markers) |
+| 5 | Layered Memory | Long-term (top 15) + Short-term (top 10) + Rolling summary |
+| 6 | Social Memory | Relationship impressions of message senders |
+
+### LLM Output → Memory Updates
+
+| # | Output | Description |
+|---|--------|-------------|
+| 1 | Reply | `messages[]` — actual messages sent to group chat |
+| 2 | Memory Ops | `memoryOps[]` — add/delete long-term & short-term entries with importance and category |
+| 3 | History Summary | `memorySummary` — compressed key points, incrementally appended |
+| 4 | Social Memory | `relationshipOps[]` — update impressions (≤200 chars) and affinity (1–100) |
+| 5 | Actions | Tool calls (shell, file ops) or command execution |
+
+### Social Memory (Relationships)
+
+Each employee maintains an impression table for every other employee (including the Boss):
+
+| Field | Description |
+|-------|-------------|
+| `employeeId` | Target employee ID |
+| `name` | Target name |
+| `impression` | ≤200 characters (e.g., "Great coder, helped me debug, reliable") |
+| `affinity` | 1–100 scale (1=hostile, 50=neutral, 100=best friend), ±5–15 per interaction |
+
+- Social memories are only injected when the corresponding employee appears in the chat
+- Affinity displayed as emoji: ❤️(≥80) 😊(≥60) 😐(≥40) 😒(≥20) 💢(<20)
+- Token cost is minimal: ~200 chars per person, 10 colleagues ≈ 1000 tokens
+
+### Memory Management Strategies
+
+| Strategy | Description |
+|----------|-------------|
+| **Incremental Ops** | AI uses `memoryOps` (add/delete/update), never full replacement — prevents accidental data loss |
+| **Dedup Check** | New entries are checked against existing first 80 characters |
+| **TTL Auto-cleanup** | Short-term memories expire automatically |
+| **Importance Pruning** | When long-term exceeds 200 entries, lowest importance entries are dropped |
+| **Rolling Compression** | When history summary grows too long, oldest summary segments are discarded |
+| **Persistence** | All memory is serialized via `Memory.serialize()` to local storage |
+
+---
+
+## Project Structure
+
+```
+IdeaCo/
+├── src/
+│   ├── app/                        # Next.js App Router + API Routes
+│   │   └── api/                    # RESTful endpoints for all features
+│   ├── components/                 # React UI components
+│   │   ├── PixelOffice.jsx         # Pixel-art virtual office
+│   │   ├── GroupChatView.jsx       # Group chat interface
+│   │   ├── AgentDetailModal.jsx    # Employee profile card
+│   │   ├── RequirementDetail.jsx   # Requirement workflow view
+│   │   └── ...
+│   ├── core/                       # Core engine
+│   │   ├── agent/                  # Agent layer (LLM communication)
+│   │   │   ├── base-agent.js       # Abstract base class
+│   │   │   ├── llm-agent/          # OpenAI-compatible API agent
+│   │   │   ├── cli-agent/          # CLI tool agent (Claude Code, Codex, CodeBuddy)
+│   │   │   └── web-agent/          # Browser-based web agent (ChatGPT Web)
+│   │   ├── employee/               # Employee layer
+│   │   │   ├── base-employee.js    # Core employee logic
+│   │   │   ├── lifecycle.js        # Autonomous behavior (poll, think, reply)
+│   │   │   ├── memory/             # Memory system (long/short-term, social)
+│   │   │   ├── secretary.js        # Secretary — boss's AI assistant
+│   │   │   ├── skills.js           # Skill management
+│   │   │   ├── knowledge.js        # Knowledge base integration
+│   │   │   └── performance.js      # Performance tracking
+│   │   ├── organization/           # Organization layer
+│   │   │   ├── company.js          # Company entity & state machine
+│   │   │   ├── department.js       # Department lifecycle
+│   │   │   ├── team.js             # Cross-department task forces
+│   │   │   ├── group-chat-loop.js  # Group chat event loop
+│   │   │   └── workforce/          # HR, talent market, role archetypes
+│   │   ├── system/                 # System services
+│   │   │   ├── audit.js            # Cost & usage auditing
+│   │   │   ├── cron.js             # Scheduled tasks
+│   │   │   └── plugin.js           # Plugin system
+│   │   ├── prompts.js              # All LLM prompt templates
+│   │   ├── requirement.js          # Requirement workflow engine
+│   │   └── workspace.js            # File workspace management
+│   ├── lib/                        # Frontend utilities
+│   │   └── store.js                # Zustand global state
+│   └── locales/                    # i18n (zh/en/ja/ko/es/de/fr)
+├── data/                           # Runtime data (auto-created)
+├── workspace/                      # Agent-produced files per department
+├── electron/                       # Electron desktop app shell
+├── Dockerfile
+├── docker-compose.yml
+└── package.json
+```
+
+---
+
+## Data Flow
+
+```
+👤 Boss
+ │
+ │  Chat message
+ ▼
+🧑‍💼 Secretary ──────────────────────────────────┐
+ │                                              │
+ │  Understands intent                          │  Simple task
+ │                                              │  → handles directly
+ ▼                                              ▼
+┌─────────────┐   ┌──────────────┐    ┌─────────────────┐
+│ Create Dept │   │ Create Team  │    │ Direct Response  │
+└──────┬──────┘   └──────┬───────┘    └─────────────────┘
+       │                 │
+       ▼                 ▼
+┌─────────────┐   ┌──────────────┐
+│  HR Hiring  │   │  Assign Req  │
+└──────┬──────┘   └──────┬───────┘
+       │                 │
+       ▼                 ▼
+┌─────────────────────────────────────────┐
+│            💬 Group Chat                 │
+│                                         │
+│  ┌─────────┐  ┌─────────┐  ┌─────────┐ │
+│  │ Emp A   │  │ Emp B   │  │ Emp C   │ │
+│  │ (poll)  │  │ (poll)  │  │ (poll)  │ │
+│  └────┬────┘  └────┬────┘  └────┬────┘ │
+│       │            │            │       │
+│       ▼            ▼            ▼       │
+│  ┌─────────────────────────────────┐    │
+│  │      🧠 Flow-of-Thought        │    │
+│  │  ┌────────┐ ┌───────┐ ┌──────┐ │    │
+│  │  │ Memory │ │ Social│ │ Anti │ │    │
+│  │  │ System │ │ Memory│ │ Spam │ │    │
+│  │  └────────┘ └───────┘ └──────┘ │    │
+│  └─────────────────────────────────┘    │
+│       │                                 │
+│       ▼                                 │
+│  Reply → Nudge Others → Chain Reaction  │
+└─────────────────────────────────────────┘
+       │
+       ▼
+  📁 Deliverables (workspace/)
+```

package/Dockerfile

@@ -0,0 +1,50 @@
+FROM node:20-alpine AS base
+
+# Install dependencies only when needed
+FROM base AS deps
+WORKDIR /app
+
+# Copy package files
+COPY package.json package-lock.json ./
+RUN npm ci --omit=dev
+
+# Build the application
+FROM base AS builder
+WORKDIR /app
+COPY --from=deps /app/node_modules ./node_modules
+COPY . .
+
+# Remove vendor submodules (not needed for production)
+RUN rm -rf vendor/
+
+# Build Next.js
+ENV NEXT_TELEMETRY_DISABLED=1
+RUN npm run build
+
+# Production image
+FROM base AS runner
+WORKDIR /app
+
+ENV NODE_ENV=production
+ENV NEXT_TELEMETRY_DISABLED=1
+
+# Create non-root user
+RUN addgroup --system --gid 1001 nodejs
+RUN adduser --system --uid 1001 nextjs
+
+# Copy built application
+COPY --from=builder /app/public ./public
+COPY --from=builder --chown=nextjs:nodejs /app/.next/standalone ./
+COPY --from=builder --chown=nextjs:nodejs /app/.next/static ./.next/static
+
+# Create data directories with proper permissions
+RUN mkdir -p data/memories data/audit workspace && chown -R nextjs:nodejs data workspace
+
+USER nextjs
+
+EXPOSE 9999
+
+ENV PORT=9999
+ENV HOSTNAME="0.0.0.0"
+
+CMD ["node", "server.js"]

package/LICENSE

@@ -0,0 +1,29 @@
+MIT License
+
+Copyright (c) 2025 Idea Unlimited (金点子无限公司)
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+---
+
+This project includes code and design patterns inspired by or derived from:
+
+- OpenClaw (https://github.com/openclaw/openclaw)
+  Copyright (c) 2025 Peter Steinberger — MIT License
+  See THIRD-PARTY-NOTICES.md for full license text.