I burnt out as a developer without noticing — so I built an AI that watches your GitHub commits for the warning signs

Hey IH,

I want to share something I built after a personal experience with burnout.

About a year ago, my GitHub commit frequency quietly dropped. My standups became vague. I was shipping less, feeling stuck, but couldn't articulate why. By the time I realised I was burnt out, I'd already lost months of momentum.

I started tracking my own GitHub activity manually — commit patterns, frequency, time of day, context switching between repos. The data told a clear story my brain had missed.

So I turned that process into a product: CodexPulse.

It connects to your GitHub and automatically:

- Generates your daily standup from your actual commits

- Scores your burnout risk based on real activity patterns

- Tracks your career health across all your repos

It's live at codexpulse.io. Free tier has 3 standups per month, Pro is $12 per month.

I'd love feedback from other indie hackers — especially:

- Are you solving the standup problem a different way?

- Do you think burnout can actually be measured from commit data?

- What signals would you add?

Happy to answer any questions about the build — React frontend, FastAPI backend, Supabase for the database, and Claude AI for the intelligence layer.