New Improvements
- Better AWS support
- Bedrock single-token support
- AMP IRSA support
What's Changed
- Copied mimir docs from robusta by @pavangudiwada in #882
- Fix tool call numbering to maintain sequential order across iterations by @pavangudiwada in #885
- ROB-1931: add support for single AWS bedrock token by @nherment in #890
- evals: add support for model-comparison and cost calculation by @aantn in #894
- fix test direction by @Sheeproid in #871
- Docs: List datasources in alphabetical order by @aantn in #899
- evals: better error reporting and other fixes by @aantn in #902
- Improve datadog evals by @aantn in #903
- fix datadog logs eval by @aantn in #904
- new runbooks prompt by @mainred in #829
- Add a test for traces due to db slowness by @Sheeproid in #907
- remove duplicate only setup by @Sheeproid in #908
- Run prerequisite for each cache toolset by @moshemorad in #909
- Add per-eval timeout to evals + small bug in eval results table by @aantn in #910
- Move eval 03 out of easy by @aantn in #912
- [ROB-1923] - added amp holmes irsa support by @Avi-Robusta in #914
- ROB-1931 bash toolset replaces cli toolsets by @nherment in #855
Full Changelog: 0.13.0...0.13.1