The llms-full.txt file provides a comprehensive, machine-readable version of Vercel's documentation optimized for large language models.

URL: https://vercel.com/docs/llms-full.txt

Use this file to give AI assistants full context about Vercel's platform, features, and best practices. This is helpful when you want an AI to understand Vercel comprehensively before answering questions or generating code.

Using llms-full.txt with AI tools

You can reference the llms-full.txt file in various AI tools:

• Claude, ChatGPT, Gemini: Paste the URL or content into your conversation
• Cursor, Windsurf: Add the URL to your project's context or rules
• Claude Code: Use the WebFetch tool to fetch the content

Markdown and agent discovery

Every documentation page is available as Markdown, and site-wide indexes help agents find pages and understand how they connect. Use the page-actions menu on any docs page to select View as Markdown or Copy page.

See Markdown and Agent Discovery for details on:

• Requesting any page as Markdown with the Accept: text/markdown header or the .md extension
• Using site-wide discovery indexes such as llms.txt and sitemap.md
• Accessing the complete cross-link graph at /docs/graph.json
• Exploring per-page .graph.md cross-link maps
• Feeding documentation to AI assistants

Site-wide discovery files

Agents can use these files to navigate or load the broader documentation set:

Install skills with a single command:

npx skillsadd

Skills.sh supports 18+ AI agents including Claude Code, GitHub Copilot, Cursor, Cline, and many others. The directory contains skills covering:

• Framework-specific guidance (React, Vue, Next.js, and more)
• Development tools (testing, deployment, documentation)
• Specialized domains (security, infrastructure, marketing)

See Agent Skills for the complete list of Vercel-provided skills, or browse the Skills.sh directory to find skills from the community.

TanStack Intent

TanStack Intent loads agent skills from installed dependencies, keeping the guidance aligned with the package version in your project. Use it when building with libraries that publish skills, such as TanStack AI, Start, and Query.

Follow the TanStack Intent setup instructions to select permitted skills and configure your coding agent. See Choosing between TanStack Intent and skills for how the two tools fit into your workflow.

CLI workflows

End-to-end workflows that show AI agents how to compose Vercel CLI commands into complete work sessions. Each workflow covers a full task from start to finish, including the reasoning between steps.

See CLI Workflows for the full list, including:

• Debugging production 500 errors
• Rolling back a production deployment
• Debugging slow Vercel Functions
• Deploying a project from the CLI

Build with an eve template

Deploy an eve template, then customize the agent for your use case: