Blogizi CLI

AI agent that turns your codebase into SEO blog posts.

Install

npm install -g blogizi

Setup

1. Go to blogizi.com and create a project
2. Copy your API key from the project settings
3. Run:

blogizi auth

Usage

Draft a post (saved as draft): blogizi draft --keyword "go chi middleware"

Draft and publish immediately: blogizi draft --keyword "go chi middleware" --publish

Publish a local .md file: blogizi publish ./blogizi-posts/go-chi-middleware.md

Upload a local .md file as a draft (same as publish, but not live): blogizi upload ./blogizi-posts/go-chi-middleware.md

Post Types

Choose the shape of the post with --type (defaults to tutorial):

blogizi draft --keyword "go chi middleware" --type tutorial # technical, code-heavy (default) blogizi draft --keyword "why I rewrote my auth" --type story # narrative journey, git history driven blogizi draft --keyword "chi vs gin" --type comparison # X vs Y, high search intent blogizi draft --keyword "v2.0 release notes" --type changelog # release notes from git log blogizi draft --keyword "why I chose postgres" --type opinion # why I chose X, lessons learned blogizi draft --keyword "7 go middleware patterns" --type listicle # numbered list format

Suggest Post Ideas

Not sure what to write about? Scan your recent git activity for post ideas:

blogizi suggest blogizi suggest --ai codex --limit 3

This reads your recent commits and project architecture and proposes titles, keywords, and post types grounded in what you've actually shipped — no dedup against posts already on your dashboard, since the CLI doesn't track which project an API key belongs to.

Draft directly from a suggestion:

blogizi draft --keyword "" --type

AI Providers

blogizi draft --keyword "x" --ai claude # default blogizi draft --keyword "x" --ai codex blogizi draft --keyword "x" --ai gemini

Requires the respective CLI to be installed and authenticated on your machine.

How it works

1. Reads local context suited to the chosen post type (codebase, git history, or architecture files)
2. Builds a type-specific, context-aware prompt
3. Shells out to your local AI CLI (claude -p, codex, etc.)
4. Parses the markdown output
5. Saves .md file locally in ./blogizi-posts/
6. Pushes to your Blogizi dashboard via API