Six single-job agents run his Threads; the writer may only use numbers from one file
A Japanese Threads creator split his account's operation into six AI agents with one deliverable each (researcher, writer, poster, fetcher, analyst, supervisor) and reports 1,181+ posts and over 10 million cumulative impressions. The writer agent may only cite numbers stored in a single 'truth file' so it cannot inflate results, the poster must confirm the published URL rather than the button click, and the fetcher records 'not retrieved' instead of zero when a fetch fails.
How it works
- Researcher agent lists topic candidates with audience and evidence
- Owner picks the theme and promise
- Writer drafts using only figures from the truth file
- Supervisor checks facts, duplicates and banned phrases, with a kill switch
- Poster publishes and records the live URL
- Fetcher logs reactions, marking misses as unretrieved, never zero
- Analyst leaves exactly one change to try next
Why it’s worth knowing
The safety design is the story: an agent forbidden from inventing numbers because its only allowed figures live in one file, and a fetcher that writes 'unknown' instead of a fake zero.
Where it’s from
Summarized from a public post on note.com. See the original ↗