Gal Be

Commented on autor3search · Nov 3, 2026

Good question. The file cannot stay. Restore rewrites only the files that were frozen at baseline. Then a separate check looks for test files that are not in the baseline. If it finds one, the experiment fails, even if the new test passes. The agent then reverts that commit. This stops an agent from adding an easier benchmark. To add your own tests, add them before you run baseline, or list them under unfreeze in the config.

Commented on Lookout · Oct 26, 2026

Hi DevHunt 👋 I built Lookout because Vercel shows Web Analytics one project at a time. Lookout puts every project on one board, grades each landing page and lists what to fix first. It's free, and the MCP server is open source. Feedback on the Scorecard and the Fix list is very welcome.

Commented on autor3search · Oct 25, 2026

Hey DevHunt 👋 I built autor3search because every "AI optimized my code" demo I saw skipped the hard part: measuring it. So the design starts from an adversarial premise. Assume the agent will eventually try to win by editing a test or a benchmark, and make that impossible. It's early but working, and every number we publish was measured on a real library. If you have a repo with benchmarks, I'd love for you to run it and tell me what it found, or what it didn't. Feedback very welcome!