Skip to main content

​
MCP server

Model Context Protocol (MCP) is an open standard that lets AI assistants securely access external data sources. Privy’s MCP server connects your AI coding assistant directly to our documentation, giving it live access to search and retrieve the exact information you need in real time.

​
Setup with Cursor

1

Open MCP settings

In Cursor, open Settings and navigate to Tools & MCP, then click Add MCP Server
2

Add Privy docs server

In the mcp.json configuration file, add:
3

Save and restart

Save the file and restart Cursor to apply the changes
4

Start building

Your AI assistant can now access Privy docs in real time. Try asking: “How do I add Privy login to my React app?”

​
Setup with Claude Desktop

1

Find your config file

Open your Claude Desktop configuration file:
  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
2

Add Privy MCP server

Add this to the mcpServers section:
3

Restart Claude

Restart Claude Desktop to load the new server
4

Verify it works

Ask Claude: “Search Privy docs for wallet creation” to test the connection

​
Agent skill

Privy publishes an Agent Skill — a structured capability file that tells AI coding agents what Privy can do, what inputs are needed, and common pitfalls to avoid. Unlike MCP (live doc search) or static files (raw doc dump), the skill gives your agent decision guidance, implementation workflows, and quick references that are automatically loaded when relevant.

​
Setup with Claude Code

1

Install the skill

Run this from your project root:
This creates .claude/skills/privy/SKILL.md in your project directory.
2

Verify it works

Open Claude Code and run /skills to confirm the Privy skill appears. Your agent now automatically loads Privy context when you work on authentication, wallets, or blockchain interactions.
3

Commit to your repo

Commit the .claude/skills/ directory so everyone on your team gets the skill automatically.
If you prefer not to use the skills CLI, you can install manually:

​
Setup with other agents

The Privy skill works with any tool that supports the Agent Skills standard, including Cursor, VS Code Copilot, Gemini CLI, OpenAI Codex, and more.
The CLI detects your installed agents and places the skill in the correct directory for each one.

​
What the skill provides

​
Static docs files

If your AI tool doesn’t support MCP yet, you can use one of Privy’s static documentation files instead.
Static files are snapshots and may not include the latest updates. Use MCP when possible for always current docs.

​
Which file should I use?

  • llms.txt is a curated overview of Privy. It is best for routing an agent to the right docs, understanding product structure, and choosing the right integration path.
  • llms-full.txt is a full static snapshot of the documentation. It is best when your tool needs more exhaustive reference material in one file.

​
Setup with Cursor

1

Open docs settings

Go to Settings > Features > Docs
2

Add the overview file

Click “Add new doc” and paste: https://docs.privy.io/llms.txt
3

Optionally add the full snapshot

If your agent needs deeper static reference coverage, add: https://docs.privy.io/llms-full.txt
4

Reference in chat

Use @docs -> Privy in your AI chat to reference the documentation

​
Setup with Claude Desktop

1

Download the overview file

Download the curated overview from: https://docs.privy.io/llms.txt
2

Optionally download the full snapshot

Download the full documentation snapshot from: https://docs.privy.io/llms-full.txt
3

Add to your project

Save one or both files in your project directory or a known location on your system
4

Reference in chat

Drag and drop the file into your Claude Desktop chat, or use the attachment button to upload it. Claude will then have access to the uploaded Privy documentation context for that conversation