Show HN: Cwhy – Explain and fix terminal errors using AI (written in Go)

Hi HN,

I built cwhy this weekend because I was tired of fixing the same Terraform and Kubernetes errors multiple times.

It’s a CLI tool (written in Go) that acts as a pipe for your stderr. Usage: `terraform apply | cwhy` or `docker logs container | cwhy`

It does two things: 1. Explains the error using OpenAI (GPT-3.5/4). 2. Caches the fix to a shared "Team Memory" (Supabase). If a teammate encounters the exact same error hash later, they get the cached solution instantly without hitting the AI API.

The stack is simple: - Go for the CLI (cobra) - OpenAI API for generation - Supabase for the shared vector/hash storage

I’m currently scrubbing standard API keys before sending logs, but looking for feedback on better PII sanitization strategies.

Repo is here: https://github.com/faalantir/cwhy Binary releases available for Mac/Linux/Windows.

Would love to hear your thoughts on the "shared memory" approach.