In this episode, I sit down with Boris, the creator of Claude Code and one of the key builders behind Claude Cowork, to unpack what Cowork actually unlocks and how people use it in the real world. He walks through a hands-on demo where Cowork organizes files, extracts receipt data, builds a clean spreadsheet, and even drives the browser to create and share a Google Sheet. We go deep on how “agentic” work feels different when the model takes actions across your computer, your browser, and your tools. Then I shift into Boris’s viral workflow for Claude Code: parallel sessions, plan-first execution, Claude.md as a compounding team memory, and verification loops that dramatically improve output quality.
Key Points
* I use Cowork as a “doer,” not a chat: it touches files, browsers, and tools directly.
* I think about productivity as parallelism: multiple tasks running while I steer outcomes.
* I treat Claude.md as compounding memory: every mistake becomes a durable rule for the team.
* I run plan-first workflows: once the plan is solid, execution gets dramatically cleaner.
* I give Claude a way to verify output (browser/tests): verification drives quality.
Numbered Section Summaries
1. Cowork Makes Claude Code Feel Like A Teammate
I frame Cowork as a UI-first way to access Claude Code that feels approachable for non-technical users. Boris and I focus on real use cases, especially the “work on my stuff” mindset around files and day-to-day operations.
2. “Agentic” Means Actions, Not Answers
Boris draws a clear line between chat-style tools and agents that take action across your computer. We talk about why tool use and computer use matter, and why that direction has been core to Anthropic’s roadmap.
3. Demo: Clean Up Your World With Folder Access
We start simple: granting access to a receipts folder and renaming files to match receipt dates. It’s an easy first workflow that helps people build intuition for how Cowork operates with real files.
4. Demo: From Receipts → Spreadsheet → Google Sheet
We push the demo into “real work”: extracting receipt details into a spreadsheet, then moving it into Google Sheets through browser control. This is where the mental model shifts—Cowork becomes an operator across apps.
5. Parallelism Beats Speed
I bring up the obvious critique (“humans can do this faster”), and Boris explains the real advantage: running multiple tasks in parallel while you bounce between them. The workflow becomes tending to multiple agents instead of doing every click yourself.
6. Skills, Extensions, MCP: When To Customize
We talk about keeping Cowork simple early, then adding skills when you hit specific software workflows. Skills act like repeatable procedures that help the agent perform better inside specialized tools.
7. Boris’s Claude Code Setup: Plan → Execute → Verify
I break down Boris’s viral thread on how he works: run many sessions, start in plan mode, then switch into execution once the plan looks right. The signature upgrade is verification—giving Claude a way to test and confirm its own output.
8. Claude md As Compounding Engineering
We close on one of the most practical systems: a shared Claude md file checked into the repo, updated constantly by the team. Boris ties this to a “never repeat feedback” loop that turns recurring review comments into durable behavior.
The #1 tool to find startup ideas/trends – https://www.ideabrowser.com/
LCA helps Fortune 500s and fast-growing startups build their future – from Warner Music to Fortnite to Dropbox. We turn ‘what if’ into reality with AI, apps, and next-gen products https://latecheckout.agency/
The Vibe Marketer – Resources for people into vibe marketing/marketing with AI: https://www.thevibemarketer.com/
FIND ME ON SOCIAL
X/Twitter: https://twitter.com/gregisenberg
Instagram: https://instagram.com/gregisenberg/
LinkedIn: https://www.linkedin.com/in/gisenberg/
FIND BORIS ON SOCIAL
X/Twitter: https://x.com/bcherny
