How I AI

Claude Cowork 101: How to automate your workday without touching code | JJ Englert (Tenex)

· Summarized

Share this episode
Anyone with this link can open it.
Share Link
Start at 0:00

Key takeaways

  • Claude Cowork is a powerful tool for creating a personalized, automated 'daily operating system' that manages email, schedules, content, and decision-making processes.
  • Building projects in Claude involves connecting folder-based workspaces with shared memory and context, enabling continuous, relevant AI assistance.
  • Connectors and skills unlock AI's ability to perform practical tasks across apps like Gmail, Slack, and Notion, transforming from suggestion engines to active workhorses.
  • Scheduled tasks and automation can streamline daily routines, such as morning debriefs, project planning, and content creation, saving time and mental effort.
  • Trust management—from drafts to full automation—is essential when integrating AI into personal and professional workflows, allowing gradual adoption based on comfort.

Summary

Introduction to Claude Cowork and Its Capabilities

The episode begins with JJ Englert introducing Claude Cowork, a desktop AI tool from Anthropic that sits between simple chat interfaces and full terminal-based coding. JJ explains how Cowork enables users, technical or not, to automate and manage their workflows online and offline, connecting various tools like email, Slack, and calendars. The significance of building a 'daily operating system'—a structured, AI-assisted workflow—comes into focus, emphasizing its utility for solo entrepreneurs and teams alike.

Getting Started with Projects and Context Building

JJ discusses foundational concepts like creating projects as folders on your computer, and how these projects serve as repositories of files and context for AI. The use of a 'brain' file— a preferences document—helps Claude understand your style, team, and needs, enabling it to generate personalized outputs. This setup facilitates building high-quality, consistent work habits and streamlining routine tasks such as writing emails, content, or managing schedules.

Leveraging Connectors and Skills for Automation

A significant portion covers setting up connectors to tools such as Gmail, Slack, Google Calendar, and Notion, which allow Claude to access and process real-time data. Creating skills, like analyzing emails or drafting messages, turns AI into a personalized assistant. JJ demonstrates how to analyze sent emails to craft a style guide, and describes advanced skills like multi-perspective review boards—sub-agent systems with various personas—to critique and improve work before sending or publishing. These techniques exemplify how AI can act as a multi-faceted feedback mechanism.

Operationalizing AI with Scheduled Tasks and Projects

JJ explains how scheduled tasks, such as a morning debrief that reviews email, Slack, and calendar data at a set time, can make daily planning more efficient. He outlines the importance of organizing workflows within projects to maintain context and shared memory, which helps AI generate relevant, project-specific results. The discussion illustrates how a well-structured AI-driven system can automate routine planning, content creation, and decision support.

Best Practices and Ethical Considerations in AI Automation

Throughout the episode, JJ emphasizes trust levels when using AI—starting with drafts, then reading emails, and gradually enabling more autonomous actions. He discusses permissions, safety, and how to handle AI misbehavior politely but firmly. JJ advocates for thoughtful integration of AI into work and personal life, highlighting the need to tailor the extent of data sharing and automation based on comfort and the task's criticality.

Sponsors

  • Tines · Start building intelligent workflows today.
  • Cursor · The best way to code with AI.

Show notes

JJ Englert leads community enablement at Tenex. In this episode, JJ provides a complete zero-to-one tutorial on Claude Cowork, Anthropic’s desktop tool that sits between simple chat and full terminal-based coding.


What you’ll learn:

  1. How to create your first Claude Cowork project by connecting a folder on your computer and building context over time
  2. The “brain” file strategy: how to create a preferences document that Claude reads every time to understand who you are and how you work
  3. Why one-click connectors to Gmail, Slack, Notion, and Google Calendar unlock AI that actually does work instead of just suggesting it
  4. How to analyze your sent emails to build a writing skill that perfectly matches your tone and style
  5. The sub-advisory-board technique: spinning up three AI agents with different personas to review your work from multiple perspectives
  6. How to set permissions for each connector so Claude only drafts (never sends) or always asks before taking action
  7. The scheduled-task workflow that creates a morning debrief by reading your email, Slack, and calendar every day at 7:30 a.m.
  8. Why projects with shared memory beat individual chat threads for consistent, high-quality AI outputs

—

Brought to you by:

Tines—Start building intelligent workflows today

Cursor—The best way to code with AI

—

In this episode, we cover:

(00:00) Introduction to JJ Englert

(02:48) What Cowork is and who it’s for

(05:49) Getting started: Opening the Cowork tab in Claude Desktop

(07:04) Understanding projects as folders on your computer

(07:54) Creating your “brain” file, with working preferences and context

(10:24) Demo: Building a daily operating system project from scratch

(12:18) How to prompt Cowork when starting a new project

(14:54) Understanding the project interface and shared memory

(18:37) Setting up connectors to Gmail, Slack, Google Calendar, and other tools

(21:00) Using connectors to analyze your emails and build personalized writing skills

(24:21) Creating a thinking-partner skill for decision support

(26:18) Cowork vs. OpenClaw

(27:18) Building a sub-advisory skill with multiple AI personas for feedback

(34:03) Advanced skill example: Multi-step newsletter creation with research and evaluation

(36:08) Setting up scheduled tasks for morning debriefs

(37:57) Going beyond one-off tasks with AI

(41:00) Progressive trust and the tradeoff of information for productivity

(44:08) Different use cases beyond work productivity

(46:08) Lightning round

—

Tools referenced:

• Claude Code: https://claude.ai/code

• Wispr Flow: https://whisperflow.ai/

• Monologue: https://www.monologue.to/

• Domo: https://www.domo.com/

• Pencil.dev: https://pencil.dev/

• Remotion: https://www.remotion.dev/

• Obsidian: https://obsidian.md/

• OpenClaw: https://openclaw.com/

• Notion: https://notion.so/

—

Other references:

• Get Started with Claude Cowork: https://support.claude.com/en/articles/13345190-get-started-with-cowork

—

Where to find JJ Englert:

YouTube: https://www.youtube.com/channel/UCv2ovDhYVtlJw4QMidLFP8Q

X: https://twitter.com/jjenglert

LinkedIn: https://www.linkedin.com/in/jj-englert-a08836a6/

—

Where to find Claire Vo:

ChatPRD: https://www.chatprd.ai/

Website: https://clairevo.com/

LinkedIn: https://www.linkedin.com/in/clairevo/

X: https://x.com/clairevo

—

Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email jordan@penname.co.

Follow this show on AudioBrevity and new episodes get summarized for you. Create a free account

Couldn't load the audio. A blocker may be stopping the show's tracking links. Open the file