How to Turn Claude Code into an Operating System with Carl Vellotti

76 snips

Mar 30, 2026

Carl Vellotti, AI/product management practitioner and educator specializing in Claude Code, shares how to stop using the tool like a chatbot and instead build skills, subagents, and an OS around context. He covers context management, subagents that preserve context, auto-triggering skills with hooks, and trustworthy Jupyter-backed data analysis. Short, practical, and focused on workflow architecture.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

ADVICE

Turn Repeated Workflows Into Skills

Turn repeated or common workflows into skills so Claude invokes standardized behavior like delegating to subagents or running checks.
Carl created a Context Guard skill that auto-recommends subagents for high-context web research tasks.

ADVICE

Have Claude Ask Clarifying Questions Automatically

Use Claude's Ask User Questions tool to have the model dynamically generate clarifying questions in a UI, avoiding vague assumptions.
Carl uses it to surface 10–12+ clarifying items and capture answers into session memory for higher-quality outputs.

ADVICE

Fix Weaknesses With Prompt-Based Skills

Improve weak areas (design, slides, research) by creating skills that either provide better prompts or wire in higher-quality tools.
Carl contrasts a plain landing page vs. using the front-end design skill which applies explicit rules like "don't look like AI."

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Today’s episode

Claude Code just hit $2.5 billion in annualized revenue in 9 months.

It is the fastest B2B software product ramp in history.

So why are most people still using it like a chatbot?

This is how most people use Claude Code. Type a prompt and get output. The context fills up. It compacts. You lose everything. You start over.

The top users flipped it. They built skills that interview through a framework before building anything. They use sub-agents that preserve context. They have operating systems where every file, every person, every project has a home.

That shift is what today’s episode is about.

I sat down with Carl Vellotti for the third time. His first episode was the beginner course. His second episode was the advanced masterclass. Together they crossed over a million views across platforms.

Today is the operating system layer. If you are already an 80 out of 100 on Claude Code, this episode will bring you to a 95 out of 100.

This episode covers context management, creating sub-agents to manage your context for you, auto-triggering skills with hooks, trustworthy data analysis with Jupyter notebooks, and building an operating system around it all.

If you are living in Claude Code 8 to 10 hours a day and want to stop fighting the tool, this is the one episode to watch.

----

Check out the conversation on Apple, Spotify, and YouTube.

Brought to you by:

* Bolt: Ship AI-powered products 10x faster

* Amplitude: The market-leader in product analytics

* Pendo: The #1 software experience management platform

* NayaOne: Airgapped cloud-agnostic sandbox

* Product Faculty: Get $550 off their #1 AI PM Certification with my link

----

If you want access to my AI tool stack - Dovetail, Arize, Linear, Descript, Reforge Build, DeepSky, Relay.app, Magic Patterns, Speechify, and Mobbin - grab Aakash’s bundle.

I’m putting on a free webinar on Behavioral and AI PM interviews. Join me.

----

Key Takeaways:

1. Context management is the real skill - A single web search eats 10% of your context. Run /context to see what is consuming it. System prompt and MCPs take 10-16% before you type one message.

2. Sub-agents save 20x context - Delegate research to a sub-agent. Same task costs 0.5% instead of 10%. Your main session only gets the summary.

3. Replace MCPs with CLIs - MCPs eat context by existing. CLIs have zero overhead. GitHub CLI, Vercel CLI, Google Workspace CLI are all dramatically more efficient.

4. Powerful skills need zero code - Anthropic's front-end design plugin is just a good prompt. No APIs or tooling. Just rules that tell Claude "do not look like AI."

5. Give Claude self-checking tools - The make slides skill uses Puppeteer to screenshot output, measure overflow, and fix issues before you see them.

6. Repeat prompts for better quality - A Google paper showed pasting a prompt twice helps. Tell Claude to double-check against skill instructions after the first pass.

7. Use hooks to auto-invoke skills - A user_prompt_submit hook matches your words against skill keywords instantly. Zero context cost.

8. Jupyter notebooks solve data trust - Every analysis shows exact code, inputs, and outputs. Traceable and reproducible.

9. Build an operating system - Knowledge folder for people context. Projects folder for task isolation. Tools folder for scripts. CLAUDE.md for identity.

10. The people folder compounds - Connect meeting transcription. After every meeting, update each person's dossier. Every prompt gets more specific over time.

----

Related content

Podcasts:

* Claude Code Masterclass with Carl Vellotti (Ep 2)

* Claude Code PM OS with Dave Killeen

* OpenClaw Setup Guide with Naman Pandey

Newsletters:

* The ultimate guide to context engineering

* How to use Claude Code like a pro

* Claude Cowork and Code setup guide

PS. Please subscribe on YouTube and follow on Apple & Spotify. It helps!

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.news.aakashg.com/subscribe