embed-ai

Embed

Embed makes your wallet, agent, or social app feel more personal with AI-powered recommendations. Deliver real-time, personalized feeds for content, swaps, and bets — directly inside your app, via one simple API. Built for low spam, high signal.

---

For AI Agents

Install the Embed skill to give your AI agent access to on-chain intelligence:

/plugin install embed-ai

Or point any LLM at our machine-readable reference:

curl https://raw.githubusercontent.com/ZKAI-Network/embed-developer/main/llms.txt

---

Links

Resource	URL
Embed website	getembed.ai
Documentation	docs.getembed.ai
Console	console.getembed.ai
SDK (npm)	@embed-ai/sdk
Pipeline SDK (npm)	mbd-studio-sdk

---

Packages

@embed-ai/sdk — Production

Serve deployed feeds, manage configs, ingest data. Single endpoint: POST /v3/for-you.

npm install @embed-ai/sdk

mbd-studio-sdk — Experimentation

Build recommendation pipelines: Search → Features → Scoring → Ranking.

npm install mbd-studio-sdk

---

Examples

• 01 — Getting Started — Polymarket recommendations demo: search → features → scoring → ranking pipeline
• 02 — Polymarket Feed — Full-stack Next.js app with personalized Polymarket prediction market feed, real-time charts, and wallet-based recommendations

---

Plugin Structure

This repo is also an installable Claude Code plugin:

├── .claude-plugin/
│   └── plugin.json              # Plugin manifest
├── skills/
│   └── embed-ai/
│       ├── SKILL.md             # What the AI agent learns
│       └── references/          # API docs, examples
├── mcp/                         # MCP server (runtime tool access for agents)
├── llms.txt                     # Machine-readable reference for any LLM
└── examples/                    # Working demos

---

License

MIT