Skip to content

Docs MCP Server

Connect Claude, Cursor, VS Code or Codex to the Pollora documentation through a remote MCP server, so your assistant searches and reads the current docs.

The Pollora documentation is available as a remote MCP server at https://pollora.dev/mcp. Once you add it to your AI assistant, the assistant can search these docs and read whole pages while it answers, instead of relying on what its training remembers about Pollora.

It needs no account and no project: it works in any editor or chat app that supports remote MCP servers, even before you install Pollora.

The two are complementary:

Docs MCP serverNectar
RunsOn pollora.devIn your project (php artisan nectar:mcp)
KnowsThe documentationYour application: routes, hooks, post types, versions
InstallAdd a URLcomposer require pollora/nectar --dev
Use it toLearn an API, find the right attribute, follow a guideWrite code that fits the project you are working on

In a Pollora project, add both.

You can also connect it from any docs page: open the menu next to Copy page.

Terminal window
claude mcp add --transport http pollora-docs https://pollora.dev/mcp

Add --scope project to save it in the project’s .mcp.json and share it with your team.

In Settings › Connectors, choose Add custom connector, name it Pollora docs and enter https://pollora.dev/mcp.

Add to Cursor in one click, or add it to ~/.cursor/mcp.json (or .cursor/mcp.json in a project):

{
"mcpServers": {
"pollora-docs": {
"url": "https://pollora.dev/mcp"
}
}
}

Add to VS Code in one click, or add it to .vscode/mcp.json:

{
"servers": {
"pollora-docs": {
"type": "http",
"url": "https://pollora.dev/mcp"
}
}
}

In ~/.codex/config.toml:

[mcp_servers.pollora-docs]
url = "https://pollora.dev/mcp"

Any client that supports remote MCP servers over Streamable HTTP can use https://pollora.dev/mcp. No authentication is needed.

The server exposes three read-only tools:

ToolWhat it does
search_docsFull-text search. Returns the best matching sections, each with its URL (down to the heading) and an excerpt.
get_pageReturns a whole page as Markdown, from its path (routing/controllers) or its URL.
list_pagesLists every page by section, with its path and summary.

Ask your assistant a question about Pollora and it calls them on its own. To make sure it does, mention it: “check the Pollora docs”.

The server answers from the documentation deployed on pollora.dev, so it follows each docs update.

Every docs page also exists as plain Markdown: add .md to its path, for example /routing/controllers.md. The Copy page button copies that version, ready to paste into a chat.

For the whole documentation in one file, use /llms-full.txt, or /llms.txt for the index.