I thought the hard part would be the scoring. Write clean YAML. Define expected behavior.
Source: [Dev.to](https://dev.to/debashish_ghosal/i-built-scenario-packs-for-agent-regression-testing-the-integration-not-the-judge-broke-me-1k9k)
I'm running a an AI fluency assessment tool and after 1800 real users who interact with our chat bot 20-40 minutes; here are some weird things we found out. 1. The least AI-fluent professionals overestimated their score by 40 points.
2 points, 0 comments on Hacker News
There are exactly two audiences that I can think of that might understand the punchlines which I am rather proud of one is in Las Vegas and the other is right here, which I can reach from my home in Texas. So please check it out. https://killsignal.
1 points, 0 comments on Hacker News
2 points, 0 comments on Hacker News
1 points, 0 comments on Hacker News