Prisma

> 数据库

开源

下一代 Node.js/TypeScript ORM

47.6K stars0 点赞6 次浏览

访问官网 GitHub

工具介绍

下一代 Node.js/TypeScript ORM

Discord | X | Blog Post | Architecture

Looking for Prisma ORM 7? It lives on the v7 branch of this repository, and the prisma and @prisma/* packages on npm continue to be published from there.

Prisma Next is currently in Early Access and we're building it in the open with the community. APIs will still evolve as your feedback shapes them, so we don't recommend it for production workloads yet. Come along for the ride: star the repo, follow @prisma on X, or read along on the Prisma blog.

Prisma Next is a TypeScript rewrite of Prisma ORM, designed to be extensible, composable, and AI-agent friendly by default. Read the full announcement: The Next Evolution of Prisma ORM.

Prerequisites

Node.js 24 or newer
A package manager (npm, pnpm, or yarn)

Getting started

1. Scaffold a new project

The interactive scaffolder picks a JavaScript framework (Next.js, Vite, Hono, and others) and wires Prisma Next in with your chosen database (PostgreSQL or MongoDB):

bash

npm create prisma@next

You finish with a runnable app, a starter contract, and the agent skills already registered.

2. Or, add Prisma Next to an existing project

Run this from your repo root:

bash

npx @prisma/cli@next orm init

prisma orm init writes prisma.config.ts, scaffolds a starter contract and db.ts under src/prisma/, installs the runtime, emits the contract, and registers the agent skills. It does not touch your framework or build setup.

3. Use your AI agent for everything Prisma Next

Both installers leave a top-level prisma-next.md primer at your project root for any agent to read first. init additionally materialises one SKILL.md per workflow in two places, so different agent runtimes can find them at their expected paths:

.claude/skills/<skill-name>/SKILL.md — picked up by Claude Code
.agents/skills/<skill-name>/SKILL.md — universal location for Cursor, Copilot Agent, and other runtimes

A skills-lock.json at the project root tracks which skill versions are installed. Your editor's AI assistant auto-loads the right skill when your prompt matches.

Just describe what you want. For example:

"Add a posts model with a relation to users, then write a query that loads each user's three most recent posts."

The agent loads the prisma-8 skill, opens its contract and queries references, then drives the change end-to-end.

For the full catalogue and what each skill covers, see skills/README.md.

Found a bug, missing a feature, or have a question for the team?

Ask your agent. The prisma-8 skill's feedback flow drafts a structured GitHub issue or hands you a Prisma Discord link for live Q&A. You can review and confirm before anything is submitted.

For extension authors

Prisma Next has a minimal core. Everything around it, including Postgres support itself, is built on the same public SPI that's available to any author. If you've wanted to integrate your tool, your database, or your library with Prisma, this is the way in.

A few extensions already shipping:

@internal/extension-pgvector: vector columns and similarity operators for semantic search.
@internal/extension-paradedb: typed BM25 indexes with multiple tokenizers for full-text search.
@internal/extension-postgis: geospatial types and queries.
@cipherstash/prisma-next: searchable encryption and data-level access control.

Want to ship your own? The Authoring Prisma Next Extensions blog walks through the SPI, the layers your extension can hook into, and how the team features new extensions in the Prisma Next directory.

Supported databases

Prisma 8 carries PostgreSQL to general availability — and that is all at this stage:

PostgreSQL — the primary target; general availability at Prisma 8
MongoDB — early access; proves the framework works beyond SQL
SQLite — a proof of concept today

MySQL follows later. See the roadmap for what must happen before the 8.0.0-rc.1 release.

Contributing

See CONTRIBUTING.md for setup, commands, DCO signoff, and PR expectations. For substantive changes, please open an issue first so we can give direction-fit feedback before you invest implementation time.

Security issues: follow the Private Vulnerability Reporting flow in SECURITY.md. Please do not file them as public issues.