Analyzer agents trimmed his agent skill files by a fifth and found bugs doing it
Analyzer agents audited and trimmed skill files, cutting about 18 percent then about 600 more lines, and extracting bash into scripts turned up bugs.
Real things everyday people got it to do, and how to try them yourself.
A random mix from the whole library. Shuffle for a new set.
Analyzer agents audited and trimmed skill files, cutting about 18 percent then about 600 more lines, and extracting bash into scripts turned up bugs.
A creator who makes hundreds of NPC entries for an application, work that takes hours, handed the archiving to the AI; it archived 780 entries in two weeks.
One user listed what his AI did over a few weeks: filed his homestead exemption back to 2024 (about a $4,000 refund on the way plus roughly $2,000 a year), switched his car insurance to Progressive (about $1,500 a year), moved him to a cheaper power plan (about $800 a year), filed 20+ class-action claims for him and his wife (about $500 expected) and applied to 15+ jobs that led to interviews. He put the total at about $9,000, with forms prepared for his approval.
Charles Summers runs his demand-generation work on the AI: 8 scheduled tasks that drive HubSpot, Search Console, Gmail and LinkedIn through Chrome, backed by 578 local sessions and about 220 memory notes. Ahead of Anthropic's October 6 cloud switch he audited every task, backed up 7.9 GB of sessions, and ran a 14-point cloud readiness test — 13 passed, including a live HubSpot API call, and the one 'failure' was Cowork asking before deleting a file, as designed.
He built a home dashboard with AI that pulls a dozen sources into one wall screen for the whole family. It shows the shared calendar, weather warnings, solar and battery levels, car charge and bin days, and it only brings things up when they matter, like bins out tonight or a frost warning.
One of his AI shortcuts had prices written out, like $10/hr and $1,200. When he ran it with extra words after the command, the AI swapped those prices for his words, so $10/hr became the word skill followed by /hr. No error, no warning. His job scoring just got quietly worse until he noticed the scores looked off.
Reverse engineered the Insta360 Studio binary with AI and wrote a shim for his AMD iGPU so the app is usable.
Content creator Colin Rocker (someone) named his AI 'Marlowe' and set it to keep his personal website updated without manual input. Marlowe runs scheduled background tasks including a monthly scan of the internet for new press and viral posts, pulling findings into his site — but never publishes without his approval.
Dan Peguine's the AI watched his fully-booked doctor's cancellation queue and booked a same-day slot when an opening appeared. The case reached X as a quote-post by Meta executive Alexandr Wang, so the amplifier is promotional even though the underlying report is a user's own.
A farmer's John Deere combine had an electronics fault for days and a paid Deere mechanic could not find it. Working with AI, he built a test suite that read the machine's CanBUS messages and the traces from his oscilloscope, and compared both against a second working combine. They found two faulty controllers and the machine runs again.
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.