Summary

Velobase Harness is an open-source AI SaaS framework designed to accelerate the path from code to revenue for AI applications. It provides developers with integrated monetization, attribution, and anti-abuse infrastructure, extracted from a high-ARR product, enabling faster financial success.

How It Works

Built on the T3 stack (Next.js 16, TypeScript, tRPC), the framework employs a modular architecture. Key components include server-side ad attribution (Google Ads, X, PropellerAds), a financial-grade affiliate engine, and flexible usage-based billing. Robust anti-abuse guardrails like Redis rate limiting and Turnstile mitigate costs. The system supports unified or split service deployments (Web, Worker, API) using BullMQ for background tasks and PostgreSQL/Redis for data. This provides a comprehensive monetization and growth stack.

Quick Start & Requirements

Prerequisites: Node.js, pnpm, Docker Desktop with Compose. Installation options include the automated Velobase Launchpad or local development (pnpm install, .env setup, pnpm docker:db:up, pnpm db:push, pnpm db:seed, pnpm dev:all). Official documentation resides within the docs/en/ directory.

Highlighted Details

• Monetization: Integrated Stripe/crypto payments, affiliate engine with ledgering, usage-based billing with credit management.
• Growth: Server-side ad attribution, email outreach with A/B testing.
• Security: Anti-abuse guardrails (Redis rate limits, Turnstile, guest chat quotas).
• Scalability: Multi-LLM AI chat, 11 BullMQ workers, flexible service splitting.

Maintenance & Community

Community engagement is fostered via a Discord server, with contribution guidelines provided. Specific maintainer or sponsorship details are not detailed in the README.

Licensing & Compatibility

Licensed under MIT, allowing unrestricted use, modification, and commercial distribution without copyleft concerns.

Limitations & Caveats

The framework's extensive feature set may introduce complexity for simpler projects. Detailed performance benchmarks are not provided. Users must adhere to defined AI agent interaction rules.