AutoBE Files

   🚀 Features

• benchmark:
• Add score detail modal with phase breakdown  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1239 (e1101)
• Pipeline phase visualization, auto-deploy, and agent…  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1243 (ada84)
• Auto-deploy benchmark results to autobe.dev  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1247 (1702b)
• interface:
• Concise function calling schema descriptions.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1231 (afbf8)
• test:
• Diagnose benchmark result.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1228 (dc3d1)
• website:
• Add native benchmark leaderboard page  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1226 (80679)
• Fix qwen meetup korea presentation  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1230 (7b012)
• Finalize qwen meetup presentation.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1233 (4573c)
• Concise qwen meetup article.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1244 (63fde)
• Replace CodeShowcase placeholder with real AutoBE-generated code  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1246 (81221)

   🐞 Bug Fixes

• agent:
• Recursive relation problem.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1253 (60218)
• Enforce implicit return type for select() function.  -  by @samchon in https://github.com/wrtnlabs/autobe/issues/1255 (8365f)
• benchmark:
• Remove gauzy project and gpt-4.1 model from benchmark data  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1236 (57d17)
• Update benchmark data with 16 models and remove gpt-4.1  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1237 (57924)
• Restore correct JSON format for benchmark-summary.json  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1241 (71ba2)
• Remove emoji icons from pipeline phase headers  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1248 (913f4)
• estimate:
• Improve evaluator accuracy and rebalance phase weights  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1229 (d19f8)
• Agent score default 100→0, add response_format json_object  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1249 (e5874)
• website:
• Remove old benchmark SPA and fix static export  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1227 (e90cf)
• Rename Benchmark to Estimate in UI text  -  by @k08200 in https://github.com/wrtnlabs/autobe/issues/1250 (34c44)

    View changes on GitHub