MCP Server
@praxisjs/mcp — Model Context Protocol server for PraxisJS. Gives any MCP-compatible AI assistant live access to documentation and install commands.
MCP Server
@praxisjs/mcp is a Model Context Protocol server for PraxisJS. It works with any MCP-compatible AI assistant — Claude Code, Cursor, VS Code/Copilot, Windsurf, and others — giving it live access to praxisjs.org documentation and install command generation.
npm install -D @praxisjs/mcppnpm add -D @praxisjs/mcpyarn add -D @praxisjs/mcpbun add -d @praxisjs/mcpNo install required. Run directly:
npx -y @praxisjs/mcppnpm dlx @praxisjs/mcpyarn dlx @praxisjs/mcpbunx @praxisjs/mcpSetup
Pick your AI client and add the server to its MCP config file.
Add to .claude/settings.json (project) or ~/.claude/settings.json (global):
{
"mcpServers": {
"praxisjs": {
"command": "npx",
"args": ["-y", "@praxisjs/mcp"]
}
}
}Add to .cursor/mcp.json at the project root (or ~/.cursor/mcp.json for global):
{
"mcpServers": {
"praxisjs": {
"command": "npx",
"args": ["-y", "@praxisjs/mcp"]
}
}
}Then enable it in Cursor Settings → MCP.
Add to .vscode/mcp.json at the project root:
{
"servers": {
"praxisjs": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@praxisjs/mcp"]
}
}
}Click Start in the config file editor, or use the MCP panel in the GitHub Copilot Chat sidebar.
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"praxisjs": {
"command": "npx",
"args": ["-y", "@praxisjs/mcp"]
}
}
}Add to ~/.codex/config.toml:
[mcp_servers.praxisjs]
command = "npx"
args = ["-y", "@praxisjs/mcp"]If you prefer a local install over npx, run npm install -D @praxisjs/mcp and replace "command": "npx", "args": ["-y", "@praxisjs/mcp"] with "command": "./node_modules/.bin/praxisjs-mcp".
Tools
The server exposes four tools.
praxisjs_overview
Returns the documentation index (llms.txt) — a concise list of all available pages with one-line descriptions. Use this first to orient before fetching a specific page.
praxisjs_get_page
Returns a specific documentation page as Markdown. Pass a slug from the table below.
| Topic | Slug |
|---|---|
| Components | essentials/components |
| Reactivity & signals | essentials/reactivity |
| JSX | essentials/jsx |
| Lifecycle hooks | essentials/lifecycle |
| Async data | essentials/async-data |
| State & Props | decorators/state |
| Watchers | decorators/watchers |
| Events & Slots | decorators/events |
| Performance | decorators/performance |
| Timing decorators | decorators/timing |
| Utility decorators | decorators/utilities |
| Router | ecosystem/router |
| Store | ecosystem/store |
| Dependency injection | ecosystem/di |
| Motion | ecosystem/motion |
| State machines | ecosystem/fsm |
| Concurrency | ecosystem/concurrency |
| DOM composables | composables/dom |
| Browser composables | composables/browser |
| Custom decorators | advanced/custom-decorators |
| Custom composables | advanced/custom-composables |
| Vite plugin | tooling/vite-plugin |
praxisjs_full_docs
Returns all documentation pages concatenated (llms-full.txt). Use when cross-topic context is needed or the relevant page slug is unknown.
praxisjs_get_install_command
Generates the correct CLI command to install @praxisjs/* packages without pinning version numbers. Use this whenever a PraxisJS package needs to be installed — never write version constraints directly into package.json.
Input:
| Field | Type | Default | Description |
|---|---|---|---|
packages | string[] | — | One or more @praxisjs/* package names |
manager | "npm" | "pnpm" | "yarn" | "bun" | "npm" | Package manager to use |
dev | boolean | true | Install as devDependency |
Examples:
packages: ["@praxisjs/router", "@praxisjs/store"]
manager: "pnpm"
→ pnpm add -D @praxisjs/router @praxisjs/storepackages: ["@praxisjs/core", "@praxisjs/runtime"]
manager: "npm"
dev: false
→ npm install @praxisjs/core @praxisjs/runtimePairing with an AI skill
The MCP server handles documentation lookup. For AI-assisted development that also enforces project memory (CLAUDE.md / AGENTS.md), one-time project config, and PraxisJS conventions, add a skill to your project:
# New project — select the integration during scaffolding
pnpm create praxisjs@latest
# Existing project
npx create-praxisjs addSee the AI skills guide for what each skill does and how to use it.