# SynaBun

> Open-source vibe coding toolkit with persistent AI memory, social media automation, and 66 MCP tools for agentic coding

SynaBun is an open-source agentic coding toolkit with 66 MCP tools. It combines persistent vector memory, social media automation, browser automation via Playwright, a visual whiteboard, Discord integration, and autonomous agent loops into a single MCP server connection. Built for vibe coding with Claude Code, Cursor, Windsurf, and any agentic IDE or MCP-compatible AI assistant. SynaBun gives AI assistants persistent memory across sessions — your AI remembers architecture decisions, bug fixes, and coding patterns instead of starting from zero every time.

## Core Capabilities

### Social Media Automation
- **Twitter/X**: Extract tweets, timelines, and engagement data as structured JSON. Compose and interact via AI-driven browser control.
- **TikTok**: Extract feed videos, search results, creator profiles, and TikTok Studio content with full metadata.
- **Facebook**: Parse posts from feeds, groups, and Pages with author info, timestamps, and reactions.
- **Instagram**: Extract feed posts, profiles, reels, search results, and individual post details with full metadata.
- **LinkedIn**: Extract feed posts, profiles, notifications, messages, people search, network connections, and job listings.
- **WhatsApp**: Read chat lists and message threads from WhatsApp Web.

### Browser Automation
- Full Playwright-powered Chromium browser with persistent sessions
- Navigate, click, fill forms, type, screenshot, extract content, execute JavaScript
- Persistent cookies and login state across AI conversations
- Headed or headless mode

### Persistent Vector Memory
- Store and recall knowledge as semantic vectors using SQLite
- Semantic search — find context by meaning, not exact keywords
- 12+ embedding providers (Transformers.js local default, OpenAI, Ollama, Gemini, Cohere, Mistral, and more)
- Multi-project support with isolated memory spaces

### Vibe Coding Features
- AI remembers architecture decisions, bug fixes, and coding patterns across sessions
- 7 Claude Code lifecycle hooks for automated memory capture and recall
- File sync detects when source files change, flagging stale memories
- Conversation memory with cross-session recall

### Visual Workspace
- Interactive whiteboard for AI collaboration and architecture diagrams
- Floating cards for pinning memories, research notes, and investigations
- 3D Neural Interface visualization of memory graphs (Three.js)
- Skills Studio for building reusable AI workflows

### Autonomous Loops
- Run long-running AI tasks with configurable intervals
- Start, stop, and monitor autonomous agent loops

## Tech Stack

- Runtime: Node.js 22.5+
- Vector DB: SQLite (local file-based, built-in via node:sqlite)
- Protocol: Model Context Protocol (MCP)
- Browser: Playwright (Chromium)
- Visualization: Three.js
- Embeddings: Transformers.js (local default)

## Quick Start

```bash
npx create-synabun@latest
```

## Links

- Website: https://synabun.ai
- Documentation: https://synabun.ai/docs
- GitHub: https://github.com/danilokhury/Synabun
- License: Apache 2.0