About This Course
Prerequisites, tools needed, and what you'll build in the Repurposing Content with AI Agents course.
TUTOR WITH THEFOCUS.AI
Copy this prompt into Claude, ChatGPT, or any external AI assistant. It points the assistant to the course instructions and links it to your student profile to track your progress and customize observations.
You are not enrolled yet. Enroll to generate a Student ID to track lesson completions and store learning notes.
About This Course
What This Course Covers
This is a hands-on tutorial that teaches you how to build a content processing pipeline using AI coding agents. You’ll convert 218 email newsletters into structured, searchable markdown with rich taxonomy metadata. Along the way, you’ll learn five powerful patterns for working with AI agents that apply far beyond email processing.
The email newsletter pipeline is the example — the patterns are the point.
Prerequisites
- A computer with macOS or Linux (Windows with WSL should work)
- Basic familiarity with the terminal
- No Python knowledge required (the AI agents write the Python)
- A newsletter archive or batch of similar files you’d like to process (optional — you can follow along conceptually)
Tools You’ll Use
| Tool | What It Does | Why |
|---|---|---|
| mise | Runtime version manager | Manages Python, Node, and tool versions from a single config file |
| Claude Code | AI coding assistant | The agent that does the research, writes code, and orchestrates subagents |
| Python 3.12 | Scripting runtime | Handles EML parsing, image downloading, and validation |
| uv | Python package manager | 10-100x faster than pip, managed through mise |
All tool installation is handled through mise — no manual Python setup required.
Output Format
The pipeline produces markdown files with YAML frontmatter. Here’s an example of a fully decorated file:
---
title: "Golden Age for Indie Devs and Engineers"
date: 2024-08-26
series: fod
episode: 64
content_type: digest
primary_topic: industry
tags: ["cursor", "lerobot", "indie-hacking", "robotics", "fine-tuning"]
people_mentioned:
- name: "Andrej Karpathy"
affiliation: "OpenAI (former)"
- name: "Andrew Ng"
affiliation: "Landing AI"
companies_mentioned: ["Cursor", "Hugging Face", "OpenAI", "Anthropic", "Meta"]
models_mentioned: ["GPT-4o", "Phi-3.5", "Hermes 3", "Jamba-1.5"]
audience: mixed
depth: overview
has_research_papers: true
has_code_examples: false
is_premium: false
source: "Turing Post"
author: "Ksenia Se"
original_subject: "FOD#64: Golden Age for Indie Devs and Engineers"
original_file: "FOD#64_ Golden Age for Indie Devs and Engineers.eml"
url: "https://www.turingpost.com/p/fod64"
---The Taxonomy Schema
The classification system captures rich metadata about each piece of content. Here’s the complete field reference:
Required Fields
| Field | Type | Values |
|---|---|---|
title | string | Cleaned article title |
date | YYYY-MM-DD | Publication date |
series | string or null | fod, topic, agentic, insights, concepts, podcast, unicorn, guestpost, or null |
content_type | string | article, digest, explainer, interview, profile, curated, announcement, sponsored, forwarded, survey |
Categorization Fields
| Field | Type | Values |
|---|---|---|
primary_topic | string | llm, agents, infrastructure, research, industry, business, ethics, hardware, multimodal |
tags | array of strings | Specific technologies, techniques, or concepts |
episode | integer | Episode number (for numbered series) |
People & Organizations
| Field | Type | Description |
|---|---|---|
people_mentioned | array of objects | Each with name, optional role, optional affiliation |
companies_mentioned | array of strings | Organizations referenced |
models_mentioned | array of strings | AI models referenced (e.g., “GPT-4”, “Llama 3”) |
Content Characteristics
| Field | Type | Values |
|---|---|---|
audience | string | technical, general, business, mixed |
depth | string | overview, intermediate, deep-dive |
has_research_papers | boolean | Contains academic paper references |
has_code_examples | boolean | Contains code snippets |
is_premium | boolean | Premium subscriber content |
Source Metadata
| Field | Type | Description |
|---|---|---|
source | string | Origin publication |
author | string | Primary author |
original_subject | string | Raw email subject line |
original_file | string | Original .eml filename |
url | string | Canonical URL |
Series Identification
The newsletters follow consistent patterns detectable from subject lines:
| Series | Pattern | series Value |
|---|---|---|
| Findings of the Day | FOD#N: | fod |
| Topic Deep Dives | Topic N: | topic |
| Agentic Series | 🦸🏻#N: | agentic |
| SF Insights | 🌁#N: | insights |
| Concept Cards | Concepts: | concepts |
| Podcast/Interviews | 🎙️ | podcast |
| Unicorn Profiles | Company deep-dive structure | unicorn |
| Guest Posts | Guest Post: | guestpost |
What You’ll Walk Away With
After completing this course, you’ll understand:
- How to use AI agents for context-aware research — getting answers specific to your stack, not generic recommendations
- How to delegate code generation to AI — describing what you want and letting agents write the implementation
- How to build compound agent pipelines — each agent layer building on the research of the last
- How to scale with parallel agents — processing hundreds of files in minutes
- How to validate and iterate — using scripts to catch edge cases and improve quality
These patterns apply to any batch content processing task: migrating documentation, organizing knowledge bases, enriching datasets, building content sites from archives.