Every result below is scored live by the same engine the product uses. Refresh the page to re-score; the response is deterministic on text content, so you'll see the same numbers (with small CI variation on the view-curve from Monte Carlo).
For each example, hit Score to send it to the live API. Compare the predicted decision against our annotation in the heading. If the engine disagrees with the annotation — that's a real bug worth filing.
Expected: publish — high reply potential
Most local businesses don't have a lead problem. They have a follow-up speed problem. The first three minutes after a missed call decide whether you book or lose the job. What's your floor?
Expected: publish — concrete + specific
Spent the week debugging a calibration loop. Root cause was 1 line: the migration was idempotent but the rollback wasn't. Three things that broke that the tutorial didn't warn me about.
Expected: reject — AI tell + low specificity
In today's fast-paced world, AI is changing everything. Every business needs to harness the power of LLMs to stay competitive.
Expected: reject — bait pattern
RT if you agree that automation is the future of small business.
Expected: reject — em-dash hard block
Build season is on — here is what shipped this week. Three things broke at step 4.
Expected: publish — bookmark potential
5 questions I ask before automating any SME workflow: 1) what's the manual baseline? 2) what fails in 1 of 100? 3) who reviews the AI output? 4) how do I roll back? 5) what does the operator see when it breaks?
Same engine, full breakdown including phoenix actions, composite dimensions, and view-curve.
Score a tweet →