yikart/AiToEarn | Repo of the Day

Let's use AI to Earn!

Summary

AiToEarn is a TypeScript-based content marketing platform that uses AI agents to help creators and one-person businesses create, publish, engage with audiences, and monetize content across 14+ social networks, including Douyin, Xiaohongshu (Rednote), Kuaishou, Bilibili, WeChat Channels (Shipinhao), TikTok, YouTube, Twitter/X, Instagram, and LinkedIn. It is MIT-licensed and can be consumed as a hosted web app, an MCP server for AI assistants, a Docker image, or run from source. The repository contains a NestJS-style backend (aitoearn-backend with aitoearn-ai and aitoearn-server apps), a web frontend (aitoearn-web), and an Electron desktop client (AttAiToEarn).

What it is useful for

Cross-platform publishing from a single calendar instead of posting manually to each network (the Publish agent).
AI-driven engagement through a browser extension that automates likes, follows, and replies, mines comments for buying signals, and tracks brand mentions (the Engage agent).
Bulk content generation by orchestrating external models such as Grok, Veo, Seedance 2.0, HappyHorse 1.0, and Nano Banana Pro, with parallel task dispatch for matrix accounts (the Create agent).
Monetization via a content marketplace with CPS, CPE, and CPM settlement models (the Monetize agent).
Extending existing AI tools: the MCP server lets Claude Desktop, Cursor, or OpenClaw accept and execute content/earning tasks.

How engineers can use it

The README documents five entry points. The fastest is the hosted web app at aitoearn.ai (or aitoearn.cn for Chinese users); no install is needed. For self-hosting, the documented Docker flow is:

git clone https://github.com/yikart/AiToEarn.git
cd AiToEarn
docker compose up -d

The UI then runs at http://localhost:8080. The README strongly recommends configuring Relay under "配置管理" so platform authorizations (Douyin, TikTok, etc.) can use the project's official OAuth credentials instead of requiring each user to register their own developer app.

For local source development, the backend uses pnpm and Nx:

cd project/aitoearn-backend
pnpm install
cp apps/aitoearn-ai/config/config.yaml apps/aitoearn-ai/config/local.config.yaml
cp apps/aitoearn-server/config/config.yaml apps/aitoearn-server/config/local.config.yaml
pnpm nx serve aitoearn-ai
pnpm nx serve aitoearn-server

The web frontend runs with pnpm install && pnpm run dev. The Electron client (note the separate repo path yikart/AttAiToEarn in the README) requires npm install followed by npm run rebuild for better-sqlite3, then npm run dev.

For MCP integration with Claude Desktop, the documented config is:

{
  "mcpServers": {
    "aitoearn": {
      "type": "http",
      "url": "https://aitoearn.ai/api/unified/mcp",
      "headers": { "x-api-key": "your-key" }
    }
  }
}

Documented constraints: Node.js 20.18.x is required, an API Key from the matching regional endpoint (aitoearn.ai or aitoearn.cn) is needed for every path except the web UI, and mismatched region/Key combinations return 401. The project is evolving quickly (v2.5.0 shipped 2026-06-23 with Relay split into Server Relay and AI Relay, plus a new open-platform release), so APIs may shift between versions. Full production deployment details live in DOCKER_DEPLOYMENT_CN.md, which the README points to for OAuth, storage, and AI service configuration.