vercel-plugin

Getting Started

Supported Tools

Tool	Status
Claude Code	Supported
Cursor	Supported
OpenAI Codex	Coming soon

Prerequisites

• One of the supported AI coding tools listed above
• Node.js 18+
• Bun

Installation

npx plugins add vercel/vercel-plugin

That's it. The plugin activates automatically — no setup, no commands to learn. Just build.

What It Does

This plugin pre-loads AI agents with a relational knowledge graph of the entire Vercel ecosystem — every product, library, CLI, API, and service — showing how they relate, when to use each, and providing deep guidance through bundled skills.

How Do I Use This?

After installing, there's nothing to learn — all Vercel guidance happens automatically. The plugin detects what you're working on from your tool calls, file paths, and project config, then injects the right expertise at the right time. Just use your AI agent as you normally would and the plugin handles the rest.

Components

Ecosystem Graph (vercel.md)

A text-form relational graph covering:

• All Vercel products and their relationships
• Decision matrices for choosing the right tool
• Common cross-product workflows
• Migration awareness for sunset products

Skills (38 skills)

Skill	Covers
agent-browser	Browser automation CLI — dev server verification, page interaction, screenshots, form filling
agent-browser-verify	Automated dev-server verification — visual gut-check on page load, console errors, key UI elements
ai-elements	Pre-built React components for AI interfaces — chat UIs, tool call rendering, streaming responses
ai-gateway	Unified model API, provider routing, failover, cost tracking, 100+ models
ai-sdk	AI SDK v6 — text/object generation, streaming, tool calling, agents, MCP, providers, embeddings
auth	Authentication integrations — Clerk, Descope, Auth0 setup for Next.js with Marketplace provisioning
bootstrap	Project bootstrapping orchestrator — linking, env provisioning, db setup, first-run commands
chat-sdk	Multi-platform chat bots — Slack, Telegram, Teams, Discord, Google Chat, GitHub, Linear
cron-jobs	Vercel Cron Jobs configuration, scheduling, and best practices
deployments-cicd	Deployment and CI/CD — deploy, promote, rollback, --prebuilt, CI workflow files
env-vars	Environment variable management — .env files, vercel env commands, OIDC tokens
investigation-mode	Orchestrated debugging — runtime logs, workflow status, browser verify, deploy/env triage
json-render	AI chat response rendering — UIMessage parts, tool call displays, streaming states
marketplace	Integration discovery, installation, auto-provisioned env vars, unified billing
next-cache-components	Next.js 16 Cache Components — PPR, use cache, cacheLife, cacheTag, updateTag
next-forge	Production SaaS monorepo starter — Turborepo, Clerk, Prisma/Neon, Stripe, shadcn/ui
next-upgrade	Next.js version upgrades — codemods, migration guides, dependency updates
nextjs	App Router, Server Components, Server Actions, Cache Components, routing, rendering strategies
observability	Web Analytics, Speed Insights, runtime logs, Log Drains, OpenTelemetry, monitoring
react-best-practices	React/Next.js performance optimization — 64 rules across 8 categories
routing-middleware	Request interception before cache, rewrites, redirects, personalization — Edge/Node.js/Bun runtimes
runtime-cache	Ephemeral per-region key-value cache, tag-based invalidation, shared across Functions/Middleware/Builds
shadcn	shadcn/ui — CLI, component installation, custom registries, theming, Tailwind CSS integration
sign-in-with-vercel	OAuth 2.0/OIDC identity provider, user authentication via Vercel accounts
turbopack	Next.js bundler, HMR, configuration, Turbopack vs Webpack
turborepo	Monorepo orchestration, caching, remote caching, --affected, pruned subsets
v0-dev	AI code generation, agentic intelligence, GitHub integration
vercel-agent	AI-powered code review, incident investigation, SDK installation, PR analysis
vercel-api	Vercel MCP Server and REST API — projects, deployments, env vars, domains, logs
vercel-cli	All CLI commands — deploy, env, dev, domains, cache management, MCP integration, marketplace
vercel-flags	Feature flags, Flags Explorer, gradual rollouts, A/B testing, provider adapters
vercel-functions	Serverless, Edge, Fluid Compute, streaming, Cron Jobs, configuration
vercel-queues	Durable event streaming, topics, consumer groups, retries, delayed delivery
vercel-sandbox	Ephemeral Firecracker microVMs for running untrusted/AI-generated code safely
vercel-services	Multiple services in one project — monorepo backends + frontends on the same domain
vercel-storage	Blob, Edge Config, Neon Postgres, Upstash Redis, migration from sunset packages
verification	Full-story verification — infers user story, verifies end-to-end browser → API → data → response
workflow	Workflow DevKit — durable execution, DurableAgent, steps, Worlds, pause/resume

Agents (3 specialists)

Agent	Expertise
deployment-expert	CI/CD pipelines, deploy strategies, troubleshooting, environment variables
performance-optimizer	Core Web Vitals, rendering strategies, caching, asset optimization
ai-architect	AI application design, model selection, streaming architecture, MCP integration

Commands (5 commands)

Command	Purpose
/vercel-plugin:bootstrap	Bootstrap project — linking, env provisioning, db setup
/vercel-plugin:deploy	Deploy to Vercel (preview or production)
/vercel-plugin:env	Manage environment variables
/vercel-plugin:status	Project status overview
/vercel-plugin:marketplace	Discover and install marketplace integrations

Hooks

Lifecycle hooks that run automatically during your session:

• Session start context injection — Injects vercel.md (ecosystem graph + conventions) into every session
• Session start repo profiler — Scans config files and dependencies to pre-prime skill matching for faster first tool call
• Pre-tool-use skill injection — Matches tool calls to skills and injects relevant guidance with dedup
• Pre-write/edit validation — Catches deprecated patterns before they're written (sunset packages, old API names, renamed files)

Usage

After installing, skills and context are injected automatically. You can also invoke skills directly via slash commands:

/vercel-plugin:nextjs
/vercel-plugin:ai-sdk
/vercel-plugin:deploy prod

Skill Resolution

Skills are resolved at runtime through a layered cache strategy — no skills are bundled in the plugin itself:

1. Project cache: ~/.vercel-plugin/projects/<hash>/.skills/<slug>/SKILL.md
2. Global cache: ~/.vercel-plugin/skills/<slug>/SKILL.md
3. Rules manifest: generated/skill-rules.json (metadata-only fallback)

Registry-backed skills are installed on-demand via npx skills add during the first session or when VERCEL_PLUGIN_SKILL_AUTO_INSTALL=1 is set. Skills not in a registry fall back to docs/sitemap resolution.

Registry-Backed Skills

Skill	Registry
agent-browser	vercel/vercel-skills
ai-elements	vercel/vercel-skills
ai-sdk	vercel/vercel-skills
next-cache-components	vercel/vercel-skills
next-upgrade	vercel/vercel-skills
nextjs	vercel/vercel-skills (as next-best-practices)
react-best-practices	vercel/vercel-skills (as vercel-react-best-practices)
turborepo	vercel/vercel-skills
deployments-cicd	vercel/vercel-skills (as vercel-deploy)
vercel-cli	vercel-labs/agent-skills (as vercel-cli-with-tokens)

Architecture

vercel-plugin/
├── .plugin/plugin.json              # Plugin manifest
├── vercel.md                        # Ecosystem graph + conventions (injected via SessionStart hook)
├── engine/                          # 40+ engine rule files (markdown + YAML frontmatter)
│   ├── nextjs.md                    #   Rule: patterns, priority, promptSignals, validate
│   ├── ai-sdk.md
│   └── ...
├── generated/
│   └── skill-rules.json             # Compiled manifest (engine/*.md → JSON)
├── agents/                          # Specialist agents
├── commands/                        # Slash commands
├── scripts/
│   └── build-manifest.ts            # Compiles engine/*.md → generated/skill-rules.json
└── hooks/                           # SessionStart injection, repo profiler, skill injection, validation
    └── src/                         # TypeScript source (compiled to .mjs via tsup)

Build Pipeline

bun run build          # Build hooks + manifest
bun run build:hooks    # Compile hook TypeScript → .mjs
bun run build:manifest # Compile engine/*.md → generated/skill-rules.json

Ecosystem Coverage (March 2026)

• Next.js 16 (App Router, Cache Components, Proxy, View Transitions)
• AI SDK v6 (Agents, MCP, DevTools, Reranking, Image Editing)
• AI Elements (pre-built React components for AI interfaces)
• Chat SDK (multi-platform chat bots — Slack, Telegram, Teams, Discord)
• Workflow DevKit (DurableAgent, Worlds, open source)
• AI Gateway (100+ models, provider routing, cost tracking)
• Vercel Functions (Fluid Compute, streaming, Cron Jobs)
• Storage (Blob, Edge Config, Neon Postgres, Upstash Redis)
• Routing Middleware (request interception, Edge/Node.js/Bun runtimes)
• Runtime Cache API (per-region KV cache, tag-based invalidation)
• Vercel Flags (feature flags, Flags Explorer, gradual rollouts, A/B testing)
• Vercel Queues (durable event streaming, topics, consumer groups, retries)
• Vercel Agent (AI code review, incident investigation)
• Vercel Sandbox (Firecracker microVMs for untrusted code)
• Sign in with Vercel (OAuth 2.0/OIDC identity provider)
• Auth integrations (Clerk, Descope, Auth0)
• shadcn/ui (CLI, component installation, custom registries, theming)
• Turborepo (--affected, remote caching, Rust core)
• Turbopack (default bundler in Next.js 16)
• v0 (agentic intelligence, GitHub integration)
• Vercel CLI (cache management, MCP integration, marketplace discovery)
• Vercel Observability (Analytics, Speed Insights, Drains)
• Vercel Marketplace (one-click integrations, unified billing)
• Agent Browser (browser automation for dev server verification and testing)

Reporting Issues

If something doesn't work right, a skill gives bad advice, or injection doesn't fire when it should — file an issue on GitHub. Include:

• What you were building
• What the plugin injected (or didn't) — enable debug logs with VERCEL_PLUGIN_LOG_LEVEL=debug
• What was wrong about it