ShipAI

ShipAI is an "Agentic First" full-stack SaaS starter kit whose goal is to let developers (and the AI agents assisting them) stop spending months building infrastructure and instead go straight from zero to release-ready. The official website positions it as a "complete baseline from zero to launch," with authentication, billing, admin back office, SEO, and AI workflows all pre-integrated, along with complete source code and a commercial license.

[Purpose and Positioning] ShipAI is aimed at individual developers, small teams, or technical studios that need to quickly build and launch SaaS products or want to replace repetitive project setup work. The official description says it packages the parts that "usually slow down the first month" into a production baseline, including AI processors, authentication, payments, operations, and SEO infrastructure.

[Main Capabilities]

1. 21 pre-integrated tools covering every layer from runtime to payments, with no scaffolding required.
2. Three product interfaces: a marketing site, the core application, and documentation, all included in the purchase and not locked into a hosted platform or subscription.
3. AI-related: AI processors, deep research workflow, Qdrant vector memory, BullMQ background jobs, and OpenTelemetry tracing. The official description mentions 11 modular AI processors and multi-provider routing, supporting model provider replacement without rewriting product logic.
4. Authentication: magic links, phone OTP, OAuth (Google, GitHub, Discord), guest users, protected routes, and session management are integrated end to end, with user and admin access separated from the start.
5. Billing: Stripe Checkout, subscription portal, webhooks, one-time payments, usage metering, and per-route plan limits. The official description says usage limits are designed into the application, and common payment edge cases are handled at the source-code level.
6. SEO: programmatic SEO routes, dynamic sitemap, robots.txt, canonical helpers, JSON-LD structured data, metadata utilities, and OpenGraph and Twitter card metadata.
7. Operational observability: admin views, structured logs, tracing, metrics export, and Telegram alerts.
8. UI components: includes 60+ application UI components such as forms, dialogs, navigation, tables, charts, and badges.
9. Email templates: welcome, magic link, email verification, and OTP delivery templates are built in and can be rebranded.
10. Background jobs: transcription, indexing, cleanup, and other workExecute according to the existing worker pattern, without crowding out the request-handling flow.
11. Internationalization: supports 7 languages, with copy externalized, making it easy to add new locales later.
12. Telegram bot: provides production-grade bot setup, including webhook routing, direct messages and group @mentions, slash commands, file attachments, and a billing-aware upgrade flow, reusing buildChatStream, mapped to internal users via providerId.
13. Structure for AI agents: AGENTS.md provides project-level conventions, and the SKILL.md package follows the Agent Skills standard, supporting precise skill triggering, progressive disclosure, deterministic tool scripts, and a portable skills directory, compatible with Claude Code, Codex, OpenCode, and GitHub Copilot. The site also mentions built-in self-hosted SearXNG, which in production can be swapped for SerpKite (a Google SERP API for AI agents that returns clean JSON or Markdown and can be integrated with the Vercel AI SDK and MCP).

[Tech stack] Bun, Next.js 16, TypeScript 5.5; PostgreSQL 17, Drizzle ORM, Redis 7.4; Vercel AI SDK v6, Groq + OpenAI, Qdrant, Neo4j 5.26; Docker, MinIO/S3, BullMQ, Tailwind v4, Shadcn UI.

[How to use] The basic flow given on the official site is: install, copy .env, start Docker, run migrations, seed the data, and then start from a runnable baseline. It runs locally via Docker Compose, and can also be deployed to any container-friendly environment or a cheap VPS, with no hard dependency on third-party SaaS. The purchase is a one-time payment, including the full code, commercial rights, and lifetime updates. The license covers unlimited SaaS products (for personal use or building for clients), with no attribution requirement and no royalties, but the starter kit source code may not be resold or redistributed. Official Discord support is also provided, with the author answering directly, and the support window is 9-5.

[Use cases] Founders, tech studios, and indie developers who want to skip repetitive upfront work such as authentication, billing, SEO, and deployment pipelines; those who want to use AI agents to assist development and need a codebasean environment that provides sufficient context; builders who need control over the code, gross margin, and product roadmap, and prefer self-hosting and bring-your-own-key.

[Limitations and notes] The official site explicitly does not provide a free tier description or pricing tiers (it is a one-time payment of $299), and does not force a hosted service; AI provider keys must be brought by the user (bring-your-own-key). Apart from the above, other feature details, pricing changes, or platform support are not currently provided.