Multi-Tool Workflows for Content Production

Producing technical content at volume is a pipeline problem. Ideation, drafting, editing, formatting, publishing. Doing each step in a different tool is how most creators work. That is also how most creators burn out. Here is the multi-tool workflow I use to publish daily without losing my weekends.

The Pipeline

Idea capture -> Outline -> Draft -> Edit -> Format -> Publish

Each step has a tool. Each tool is the simplest thing that works for that step. The tools hand off cleanly so I do not lose time copy-pasting.

Step 1: Idea Capture (Obsidian)

I keep a daily note in Obsidian. Every time I solve a problem, I write one sentence about it. Thirty minutes of work becomes one bullet in the daily log. At the end of the month, I have 30 bullets that are article candidates.

Step 2: Outline (Claude Code)

I paste the bullet into Claude Code with a prompt: 'Outline a 500-word technical article about this. Include a business problem intro, three main sections, and a concrete code example in the middle.' I get back a skeleton in 30 seconds.

Step 3: Draft (Claude Code + My Brain)

I expand the outline manually. The AI wrote the skeleton; I write the sentences. This is where my voice comes in, my specific stories, my actual opinions. No AI can replicate my voice perfectly, and frankly I do not want it to.

Step 4: Edit (Built-in spell check + Grammarly)

I run the draft through Grammarly. Fix typos, tighten sentences, check tone. Grammarly is not AI writing; it is AI editing. Totally different problem, totally different quality level.

Step 5: Format (VS Code)

Markdown with front matter. I have a snippet that drops the front matter template:

---
title: ...
slug: ...
publishedAt: ...
categories: [Technical, ...]
---

Paste the draft, fill the front matter, done.

Step 6: Publish (Python script)

A Python script POSTs the file to my CMS API. It reads the markdown, parses the front matter, validates with Pydantic, submits. Twenty lines of code. Runs in three seconds.

import frontmatter
import urllib.request
import json
 
post = frontmatter.load('article.md')
data = {
    'title': post['title'],
    'slug': post['slug'],
    'body': post.content,
    'categories': post['categories'],
    'publishedAt': post['publishedAt'],
    'status': 'published'
}
req = urllib.request.Request(
    'https://api.example.com/posts',
    data=json.dumps(data).encode(),
    headers={'Content-Type': 'application/json'},
    method='POST'
)
urllib.request.urlopen(req)

That is the whole publisher.

Why Multi-Tool

Each tool is best in class at one step. Obsidian is best for capture. Claude Code is best for outlining. VS Code is best for markdown. Forcing all of this through one 'writing platform' means everything is mediocre.

The glue between tools is plain text and simple scripts. Nothing fancy. Nothing fragile.

Time Breakdown

• Idea capture: 1 minute a day (bullets in the log)
• Outline + draft: 25 minutes per article
• Edit: 5 minutes
• Format + publish: 2 minutes

Total: 33 minutes per article. One article a day means 33 minutes a day on content. Two weeks of writing becomes 300 articles a year.

The Skill Soup Connection

This workflow is four skills combined: note-taking, AI prompting, Python scripting, and editorial judgment. None of them alone produce the outcome. Together they produce an article every day without burnout.