v0.30.5

   🚀 Features

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

   🐞 Bug Fixes

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

    View changes on GitHub