Working Effectively with Claude Code

Please log in to access your purchased courses.

This members-only course captures Phillip’s insights and tips for working effectively with Claude Code.

AI coding agent use is exploding, but embedded engineers still seem to be quite hesitant about them. Some of this is understandable. Hardware interactions aren’t always handled well. Perhaps your industry is more stringent, and “mostly works” isn’t an option. Nonetheless, there is still a lot of value in using these tools, even if they cannot be directly applied to generating large chunks of embedded code. There are skills that need to be learned in order to effectively work with these tools. There are capabilities that you can build up to get work done more effectively.

I’ve been working with Claude Code pretty extensively for several months, with great success. I’ve also been fielding questions from some of my readers and mentees about working with Claude Code effectively. I wanted to capture those lessons, tips, and resources in a common place to share with our members. Rather than wading through the great deluge of information available, start here, with stuff I’ve already applied.

This course will be regularly updated as I continue to learn more.

The Focus

I primarily use Claude Code from the command line. I also use the Anthropic APIs for some of the supplemental tools I’ve developed. The information presented here is primarily organized with that usage in mind.

A large portion of the content will be applicable with the desktop app as well, though I will not provide instructions for that at this time.

What About Other Tools?

While I primarily use Claude Code now, much of the information contained also be useful with other AI tools. Particularly these modules:

  • Putting Claude to Work – these describe general ways of working with a coding agent beyond production code generation
  • Effective Working Sessions – the general strategies apply, though specific command references like /clear will not
  • CLAUDE.md – other tools might call this AGENT.md or have a platform-specific name. It’s the same idea of giving general guidance to the agent.
  • Creating and Using Skills – skills are an open standard at this point
  • Subagents – even if there isn’t the same subagent structure, some of the concepts are portable

Putting Claude to Work

Coding agents are great at some things, not so good at others. There are many ways they can improve our workflow without directly generating production code.

Lessons

Putting Claude to Work – Introduction Debug Assistant Research and Planning Assistant

Effective Working Sessions

This section looks at different ways to optimize your working sessions in order to get more of what you want at a lower cost.

Lessons

Effective Working Sessions – Introduction Start with Plan Mode Actively Manage Session Context Capture Lessons and Insights as You Work Use the Right Model for the Job Reduce the Scope of Work Create and Reference Checklists

Creating and Using Skills

Lessons

Use Illustrative Examples in Skills Use Progressive Disclosure in Skills