Four constraints are stated plainly and all four will shape how you use it: blogs must be created by hand in the Blogger web interface, post search runs client-side by fetching and filtering, labels are derived from posts rather than managed, and API-key authentication limits which blogs are reachable at all. Knowing that before you wire it up is worth more than any feature list.
A server over Google's Blogger API for models that need to work with an existing blog. It covers blogs, posts and labels; what it cannot do is create a blog, because the Blogger API itself does not offer that.
- Your Blogger blogs listed and retrieved
- Posts listed, searched, retrieved, created, updated and deleted on a blog you name by id
- Labels listed and retrieved, assembled from the posts themselves since the API exposes no direct label endpoints
- Post search implemented on the client side — the API has no search endpoint, so posts are fetched and then filtered
A Google Blogger API key in `BLOGGER_API_KEY`, created in the Google Cloud Console with the Blogger API v3 enabled. Node.js version 16 or higher. The package is `@mcproadev/blogger-mcp-server` (1.0.4), run over stdio with `npx -y @mcproadev/blogger-mcp-server`, or installed globally with npm. Authentication is API-key only, which limits access to public blogs and to blogs you have explicitly granted access to. A Dockerfile and a Vercel configuration ship with the repository for hosted deployment.
One command plus a key — npm install -g @mcproadev/blogger-mcp-server, then supply credentials
