Descriptions:
Host Claire Vo of How I AI brings in JJ Englert, a power user at Tenex, for a zero-to-one tutorial on Claude Co-work aimed specifically at non-engineers. The episode focuses on how Co-work — Anthropic’s GUI-based agent environment — has matured into a genuinely useful daily productivity tool for people who aren’t comfortable working in a terminal or with Claude Code directly.
Englert walks through the connector system in detail, showing how one-click integrations with Gmail, Slack, and Notion allow Co-work to both take actions (drafting emails, writing docs) and ingest personal data to improve agent behavior over time. A standout demonstration has Co-work scan 30 days of sent Gmail to extract writing patterns and generate a reusable email-style skill — effectively teaching the agent to write in the user’s own voice without any manual prompt engineering.
The tutorial also covers Co-work’s sub-agent and project systems, showing how multiple parallel agents can each adopt a distinct persona (a boss, an engineering partner, a customer) to provide multi-perspective feedback on documents like PRDs. Englert describes building a ‘daily operating system’ around an anti-to-do list — recurring tasks like email drafting and calendar conflict resolution that are permanently delegated to skills. The episode is particularly valuable for product managers, solo founders, and operations professionals looking to adopt agentic workflows without writing code, using Claude Co-work as their primary interface.
📺 Source: How I AI · Published April 13, 2026
🏷️ Format: Tutorial Demo
