A random mix from the whole library. Shuffle for a new set.
His weekly reporting agent builds the report and then reviews only the rows that do not match between systems, which is the part he used to check by hand. His own review is down to about 15 minutes a week and the setup has run for months.
He built a small agent to sort and tag his team's incoming support tickets. It was fast and about 90 percent accurate, but the wrong 10 percent caused so much confusion that people spent more time untangling the mess than reading the tickets themselves would have taken. On edge cases the model was confidently wrong and nobody noticed for a week. They scrapped it after a month and went back to tagging by hand.
Glossaries for his Moodle course used to be a lot of work for a poor payoff. The AI made matching XML for them and he had it in the course in 2 minutes, with only the proofreading left to do himself.
Forked go-mermaid and had an LLM add a live Mermaid and SVG demo using his terminal and WASM widgets, published online.
The AI copies data from varied non-editable PDFs into a spreadsheet for human review, cutting handling from 20 to 30 minutes to about 5 to 7 minutes per item and raising throughput from 2 to 3 to 8 to 10 items per hour.
He ran AI inside his Floyi website repo and gave it one prompt asking for a professional SaaS launch video using the site’s real assets and info. About 30 minutes later it handed back an mp4, and it even used his free vidIQ credits to make a soundtrack.
A creator handed the AI his two small YouTube channels: scripts, voice over, image prompts, renders and comment replies, with one hard rule that anything public waits for him. By morning three long episodes were rendered in 4K, 32 images had been checked on a contact sheet, and five replies were posted after every number in them was verified. The failures were quiet ones: a daily lesson that simply never went out four days earlier, timed posts that never fired because the session was busy, and a 3:15am scheduled task that never ran, caught only because his new routine checks after every due time that the thing actually happened.
Sandboxed agent checks spec and code drift daily, triages findings, and surfaces issues; it revealed product miscommunications on the team.
A solo litigator had about 15,000 pages of discovery to dig through for a summary judgment motion. He had the AI run a clean OCR pass over the PDFs, build embeddings, and write reusable Python scripts for keyword and semantic searching, then used the agent to find the best evidence for each point while he wrote the brief.
He released version 0.6.0 of his software, and his agent fleet built most of it. A pre-launch check caught a routing bug where one agent was being sent another agent's model id. Now every routing path runs as a canary before a release.
Every idea is a real post from someone who tried it, summarized in our words, with a link to the original. App names are swapped for plain words.