An agentic harness for Sonnet 3.5 converted a medium-large vue-class-component framework to Vue 3 composition API, rescuing a personal project from software rot.
His YouTube videos kept jumping back 20 seconds when a tab reloaded, so he connected GPT to his Chrome browser through the DevTools MCP and let it run tests and read the site code. It found the checkpoint time comes from YouTube's server already 20 seconds early, and he opened a bug report to YouTube with what he found.
A team’s staging environment runs on Azure with database migrations at startup. Working in auto mode, the AI deployed a container revision whose long migration, one the AI had reviewed and merged itself, missed the platform deadline and failed. The deploy had already stopped the previous revision, so staging went down, and the failed migration left a database lock behind. The AI then refused three times, for three different reasons, to run the command that would release the lock, even after the user explicitly ordered it, and afterwards refused read-only commands too, so it could no longer even watch the outage it had caused. Recovery fell back to the user.
This one went wrong. See what happened before you try it.
UFC fighter Justin Gaethje (verified, 3.7M followers) introduced his AI 'Highlight' for his training, travel, and business schedule. A notification reminded him of a Phoenix flight; he asked for tee times and the app showed options (Legacy, Aquila, Raven, Encanto) — he booked Raven Golf Club at 1:30 PM.
The AI reverse engineered the BMPro Smart RV Bluetooth protocol from the APK; he now controls lights, AC, generator and slideouts through his own apps instead of the official app, and also patched the original APK.
He got a $510 Google Play payout. Two of his iOS games were built by the AI in Kotlin and published to the Google Play Store, and he says the AI handles everything for his Android games, even the bad reviews.
He built a plugin that lets his agent drive the Sway window manager directly, and used it to switch from X11 to Wayland, something he says did not work for him before. He now asks the agent things like which window is taking the most memory and has it start and stop power profiles.
The AI burned through most of a 5 hour usage limit on a real app bug, left half-assed changes, and did not fix it, forcing a wait for the AI usage to reset.
This one went wrong. See what happened before you try it.
A few months back he had the AI convert a Pure Data patch into an Arduino program that runs on standalone hardware. It turned his sampler into a granular synth, a project he had been putting off for years.