Claude Therapy Tracker: using repos for more than just code

I've been using git repositories for all sorts of non-code purposes — writing white papers, managing data projects, gathering movie recommendations, and more. The latest addition to that list is therapy tracking. I put together Claude Therapy Tracker as a template for anyone who wants to use Claude Code to organize their mental health journey.

Why a repo for therapy?

The idea might sound unusual, but the utility of robust version control and organization extends far beyond code. Therapy involves tracking progress over time, documenting sessions, setting and revisiting goals, and reflecting on patterns. All of these benefit from the same structured approach that makes repos great for software projects.

What I've found matters most isn't the specific use-case — it's creating dedicated spaces for Claude to work within. When you give an AI agent a well-organized workspace with clear directories and context files, the quality of its output improves dramatically.

How it's structured

The template separates human input from AI output at the top level. There are dedicated directories for therapy session documentation (with sub-folders for pre-session plans and post-session debriefs), goal tracking, long-term planning, and AI-generated suggestions. A context/ directory serves as lightweight in-repo RAG, storing background information that helps Claude provide personalized support.

The AI infrastructure includes specialized subagents for specific tasks like logging sessions, managing context, and planning. Slash commands provide quick access to common workflows, so you can quickly debrief after a session or review your goal progress without setting things up from scratch each time.

The broader pattern

This template follows a system I've developed for working with AI on projects. The key insight is that the repo structure itself becomes part of the prompt — directories and file organization tell Claude Code what to expect and where to put things. Whether you're tracking therapy, managing a research project, or organizing personal goals, the same principles apply: separate inputs from outputs, provide rich context, and use version control to track how things evolve.

Check out the repo on GitHub if you're interested in trying this pattern for your own non-code workflows.