About

I vibe coded this
on a Friday night

I always end up in deep dives with ChatGPT and Gemini around topics I do not fully understand yet, but I almost never go back and reread those conversations.

ross.wtf exists to turn those messy threads into something worth revisiting: interactive articles you can learn from, and audio you can listen to later.

The setup is simple: one agent writes and enriches the article with custom React learning components, and another agent turns the finished piece into a narrated ElevenLabs MP3 and compresses it.

Two-Agent Workflow
01

Deep Dive

I go deep with ChatGPT and Gemini on things I do not fully understand yet (or just want to explore). The problem: I rarely reread those threads, and they get hard to follow once questions pile up.

02

Article Agent

I can pass a past LLM conversation to one agent and it turns that raw thread into a readable article, then sprinkles in interactive learning elements with custom React components.

03

Narration Agent

Once the article is finished, a second agent creates a narration script and calls ElevenLabs to read it aloud.

04

Audio Optimize

After generation, that same workflow optimizes and compresses the MP3 so each post can be explored by reading, clicking, or listening.