Step 01
Choose a Complex Topic
Choose a topic broad enough to explore but narrow enough to organize. Write down several questions you genuinely want answered so the site has a purpose beyond summarizing the field.
BeginnerPersonal
How to Build a Personal AI-Generated Learning Center on Any Topic
Use Claude Code or Codex to build a private, interlinked learning site for a subject you want to understand. Choose the audience and depth, generate a map of the field, and keep revising the structure as your questions become more specific.
Try this with your AI assistant
Download the skill for this workflow. Version 1.0.1.
Source episode · 29:00How I AI: John Kim's Playbook for AI Transformation with Quests, Skills, and 'AI Gods'
with Claire Vo
John opens a personal neuroscience learning center generated by a coding agent, with linked sections for researchers, disorders, neuromodulators, and related topics.
Before you start
What you need
- A subject and the specific questions you want the site to help answer
- Your current knowledge level and preferred learning style
- Any trusted books, papers, courses, or websites you want included
What you’ll make
A local, navigable learning site with a map of the subject, linked concept pages, suggested learning paths, and source notes.
Tools used
- ClaudeVisitClaude
Anthropic AI assistant
- CodexVisitCodex
OpenAI's cloud-based AI software engineering agent that can execute code, run tests, and handle complex multi-file tasks autonomously.
The workflow
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
5 steps
Step 02
Select a Powerful AI Model
Use a large language model known for its coding and structured data capabilities, such as Claude Code or Codex.
Step 03
Craft a Detailed Prompt
Give the coding agent your audience, questions, preferred format, and source expectations. Ask for an interconnected curriculum rather than a pile of standalone summaries.
Example prompt
Build a private learning website about [topic] for a learner at [current level]. I want it to help me answer these questions: - [question one] - [question two] - [question three] Start with a map of the field and a recommended learning path. Create linked pages for foundational concepts, major people or schools of thought, applications, and unresolved questions. Each page should explain prerequisites and link to what I should read next. Use these trusted sources when relevant: [sources]. Add source notes for factual claims and label anything uncertain. The site should work locally and remain readable offline.
Be specific about the output format you want, such as an interlinked set of documents or a full website structure.
Step 04
Generate the Learning Center
Let the agent generate the site, then have it check navigation, internal links, and missing prerequisites before you begin reading.
Step 05
Explore Your Custom Curriculum
Use the site until you hit a confusing or shallow section. Ask the agent to restructure that area, add examples, or create a more focused path. The curriculum should change as your understanding grows.
Example prompt
Based on the website content you just generated about [topic, e.g., neuroscience], create a 7-day study plan for a beginner. For each day, identify a core concept, list 2-3 key pages from the site to read, and suggest one thought experiment to help solidify the learning.
What good looks like
- The home page explains where a beginner should start.
- Concept pages link to prerequisites and related topics without dead ends.
- The curriculum moves from foundational concepts to the questions you care about.
- Claims that need support include a source or are clearly marked for further research.
Build your next product with ChatPRD
Turn an idea into a PRD, user stories, and a plan.
Runbook notes
How to recover when the loop fails and where human judgment helps.
If it goes sideways
- The site produces many short pages but never develops a useful learning path.
- Ask the model to identify prerequisites, merge overlapping pages, and create a recommended sequence built around five concrete questions.
- The material sounds authoritative but mixes established facts with uncertain claims.
- Require source notes for factual claims, label unresolved disagreements, and add a page listing topics that need stronger references.
- The generated site is hard to navigate or contains broken links.
- Ask the agent to crawl every internal link, repair orphaned pages, and generate an index grouped by prerequisite level.
Related workflows
intermediateProductCodexClaude
Build a Personal PM Operating System with a Shared Knowledge Repo
Create a repository-based system to provide an AI assistant with durable context about your team, projects, and communication channels. This enables it to perform routine tasks like summarizing daily updates and identifying stakeholders, acting as a personalized chief of staff.
Open
intermediateProductMeta Muse
Create a Personalized Family Newspaper with Meta's Muse
Use Meta's Muse to generate a printable, multi-page family newspaper that includes a weekly schedule, weather forecasts, and activity ideas. This workflow leverages Muse's strength in creating visually appealing, family-oriented artifacts.
Open
advancedProductCursorSlack
Build an Automated Growth Experiment Loop with Multiple AI Agents
Coordinate agents in Slack and Cursor to propose growth experiments for human review. Ezra’s Auto XP system was still being developed at the time of the episode.
Open
Start shipping better products.
Join 250,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.
Free to start · 14-day Pro trial · SOC 2 Type II
