v0.5.29

What's Changed

• fix connection by @jdchawla29 in #353
• Propagate ctx.reward to Trace returned by run_single_task by @shfunc in #352
• Surface per-scenario tool configs in lock file by @shfunc in #354
• Add allowed_tools rescue list to @env.scenario decorator by @shfunc in #355
• Fix job linking and taskset lookup by @jdchawla29 in #356

Full Changelog: v0.5.28...v0.5.29

v0.5.28 - CLI changes, hud dev proxy fix, tool filtering enhancements

What's Changed

• Maintenance by @jdchawla29 in #343
• Add metadata field and score alias to SubScore + tests by @shfunc in #345
• Dylanbowman314/fix tool prefixing by @dylanbowman314 in #347
• Dylanbowman314/exclude tool sources by @dylanbowman314 in #348
• patch hud dev proxy to call _hud_submit by @jdchawla29 in #349
• hotfix by @dylanbowman314 in #350
• More Maintenance by @jdchawla29 in #351

Full Changelog: v0.5.27...v0.5.28

v0.5.27

What's Changed

• small additions by @jdchawla29 in #344

Full Changelog: v0.5.26...v0.5.27

v0.5.26

What's Changed

• scenario.task by @jdchawla29 in #336
• claude enhancements by @jdchawla29 in #337
• Add annotation field to MCPToolCall by @shfunc in #339
• Task Slugs etc. by @jdchawla29 in #341
• Surface scenario evaluate errors by @farrelmahaztra in #338
• run jobs on existing tasksets only by @jdchawla29 in #342

Full Changelog: v0.5.25...v0.5.26

v0.5.25 - update bash tool, integration agent, ckpt configs

What's Changed

• Dylanbowman314/heredoc fix for other bash by @dylanbowman314 in #332
• stuff by @jdchawla29 in #333
• L/checkpoint configs by @lorenss-m in #334

Full Changelog: v0.5.24...v0.5.25

v0.5.24 - Reliability improvements for native tools

version bump

v0.5.23 - Updates to tools and context

v bump

v0.5.22

bump version

v0.5.21 - Environment conversion and robustness

Merge pull request #320 from hud-evals/l/small-fixes-3

L/small fixes 3

v0.5.20 - Backward Compatibility for Legacy Tool Spec

What's Changed

• L/small fixes 3 by @lorenss-m in #317
• J/tool compatibility by @jdchawla29 in #318

Full Changelog: v0.5.19...v0.5.20