Changelog

​

What’s Changed

• ci: skip tests in ci by @hughcrt in https://github.com/lunary-ai/lunary/pull/1168
• ci: fix release version in images by @hughcrt in https://github.com/lunary-ai/lunary/pull/1169

​

What’s Changed

• feat: add support for gpt 5.2 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1164
• chore: bump next by @hughcrt in https://github.com/lunary-ai/lunary/pull/1166
• ci: allow to deploy releases by @hughcrt in https://github.com/lunary-ai/lunary/pull/1167

​

What’s Changed

• feat: add more logging on start by @hughcrt in https://github.com/lunary-ai/lunary/pull/1163
• feat: improve refresh cost job by @hughcrt in https://github.com/lunary-ai/lunary/pull/1162

​

What’s Changed

• feat: add support for gpt 5.1 codex max by @hughcrt in https://github.com/lunary-ai/lunary/pull/1159
• feat: better support for openrouter costs by @hughcrt in https://github.com/lunary-ai/lunary/pull/1160
• fix: backend start self hosted by @hughcrt in https://github.com/lunary-ai/lunary/pull/1161

​

What’s Changed

• fix: self hosted start by @hughcrt in https://github.com/lunary-ai/lunary/pull/1158

​

What’s Changed

• feat: remove github popup by @hughcrt in https://github.com/lunary-ai/lunary/pull/1156
• fix: next vulneability by @hughcrt in https://github.com/lunary-ai/lunary/pull/1157

​

What’s Changed

• AI filters
• Many additional filter
• Sidebar revamp
• Backend performance improvements
• Analytics page performance improvements
• A brand new Dataset, Evaluations and Experiments features. This is under a feature flag and will be released little by little.

​

All updates

• feat: upadte billing delinquent banner by @hughcrt in https://github.com/lunary-ai/lunary/pull/1102
• feat: better tool rendering by @hughcrt in https://github.com/lunary-ai/lunary/pull/1104
• feat: add message count in datatable by @hughcrt in https://github.com/lunary-ai/lunary/pull/1105
• feat: dataset revamp by @hughcrt in https://github.com/lunary-ai/lunary/pull/1106
• fix: blurred conversations by @hughcrt in https://github.com/lunary-ai/lunary/pull/1107
• fix: drawer size by @hughcrt in https://github.com/lunary-ai/lunary/pull/1108
• ci: cache playwright by @hughcrt in https://github.com/lunary-ai/lunary/pull/1109
• chore: clean dependencies by @hughcrt in https://github.com/lunary-ai/lunary/pull/1110
• ci: faster test action by @hughcrt in https://github.com/lunary-ai/lunary/pull/1111
• fix: colors by @hughcrt in https://github.com/lunary-ai/lunary/pull/1112
• feat: optimize realtime evaluator job performances by @hughcrt in https://github.com/lunary-ai/lunary/pull/1113
• feat: add user intent evaluator by @hughcrt in https://github.com/lunary-ai/lunary/pull/1114
• feat: sidebar reorganization by @hughcrt in https://github.com/lunary-ai/lunary/pull/1115
• feat: improve intent evaluator by @hughcrt in https://github.com/lunary-ai/lunary/pull/1116
• feat: analytics performance improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1117
• feat: prompt sidebar ux improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1118
• feat: add datasets versions by @hughcrt in https://github.com/lunary-ai/lunary/pull/1119
• feat: reorganize evaluators by @hughcrt in https://github.com/lunary-ai/lunary/pull/1121
• feat: add retrievers and tools to log page by @hughcrt in https://github.com/lunary-ai/lunary/pull/1122
• feat: user page improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1123
• feat: prompt model config improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1124
• feat: improve datasets by @hughcrt in https://github.com/lunary-ai/lunary/pull/1125
• feat: misc by @hughcrt in https://github.com/lunary-ai/lunary/pull/1126
• chore: switch to bun for passwords hashing by @hughcrt in https://github.com/lunary-ai/lunary/pull/1127
• fix: top agents chart by @hughcrt in https://github.com/lunary-ai/lunary/pull/1128
• feat: add more chart customization by @hughcrt in https://github.com/lunary-ai/lunary/pull/1129
• feat: fix chart filters by @hughcrt in https://github.com/lunary-ai/lunary/pull/1130
• feat: evals and datasets improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1131
• feat: improve user filter by @hughcrt in https://github.com/lunary-ai/lunary/pull/1134
• chore: bump ai by @hughcrt in https://github.com/lunary-ai/lunary/pull/1135
• feat: add support for gpt 5.1 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1136
• feat: small ux tweaks by @hughcrt in https://github.com/lunary-ai/lunary/pull/1137
• chore: remove unused evaluator by @hughcrt in https://github.com/lunary-ai/lunary/pull/1139
• feat: add gemini 3 pro preview by @hughcrt in https://github.com/lunary-ai/lunary/pull/1140
• fix: anthropic tokenizer by @hughcrt in https://github.com/lunary-ai/lunary/pull/1141
• fix: anthropic tokenizer by @hughcrt in https://github.com/lunary-ai/lunary/pull/1142
• ci: add tokenizer deploy by @hughcrt in https://github.com/lunary-ai/lunary/pull/1143
• feat: add event loop lag treshold monitoring by @hughcrt in https://github.com/lunary-ai/lunary/pull/1145
• feat: add bark notification by @hughcrt in https://github.com/lunary-ai/lunary/pull/1146
• feat: log event loop blocking parts by @hughcrt in https://github.com/lunary-ai/lunary/pull/1147
• feat: simplify tokenization by @hughcrt in https://github.com/lunary-ai/lunary/pull/1148
• feat: make tiktoken encoding in workers by @hughcrt in https://github.com/lunary-ai/lunary/pull/1149
• fix: count tokens by @hughcrt in https://github.com/lunary-ai/lunary/pull/1150
• fix: token worker by @hughcrt in https://github.com/lunary-ai/lunary/pull/1151
• fix: worker path by @hughcrt in https://github.com/lunary-ai/lunary/pull/1152
• feat: more duration logs by @hughcrt in https://github.com/lunary-ai/lunary/pull/1153
• feat: add support for Claude Opus 4.5 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1154

​

What’s Changed

• feat: more aggressive rate limit for auth by @hughcrt in https://github.com/lunary-ai/lunary/pull/1099
• feat: add realtime evals verbose mode by @hughcrt in https://github.com/lunary-ai/lunary/pull/1100
• fix: realtime evaluators job by @hughcrt in https://github.com/lunary-ai/lunary/pull/1101

​

What’s Changed

• chore: bump mantine by @hughcrt in https://github.com/lunary-ai/lunary/pull/1094
• chore: bump frontend dependencies by @hughcrt in https://github.com/lunary-ai/lunary/pull/1095
• chore: bump dependencies by @hughcrt in https://github.com/lunary-ai/lunary/pull/1096
• feat: billing delinquent banner by @hughcrt in https://github.com/lunary-ai/lunary/pull/1097
• fix: prompts playground error by @hughcrt in https://github.com/lunary-ai/lunary/pull/1098

​

What’s Changed

• feat: add more rows in org top models by @hughcrt in https://github.com/lunary-ai/lunary/pull/1082
• fix: tests by @hughcrt in https://github.com/lunary-ai/lunary/pull/1085
• feat: add support for Claude Haiku 4.5 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1083
• feat: ask ai by @hughcrt in https://github.com/lunary-ai/lunary/pull/1087
• feat: better alerts by @hughcrt in https://github.com/lunary-ai/lunary/pull/1088
• fix: online evaluations results by @hughcrt in https://github.com/lunary-ai/lunary/pull/1089
• feat: improve ai filter by @hughcrt in https://github.com/lunary-ai/lunary/pull/1090
• feat: improve dropdown style by @hughcrt in https://github.com/lunary-ai/lunary/pull/1091
• feat: improve filter serialization by @hughcrt in https://github.com/lunary-ai/lunary/pull/1092
• feat: upgrade default eval model by @hughcrt in https://github.com/lunary-ai/lunary/pull/1093

​

What’s Changed

• fix: alerts by @hughcrt in https://github.com/lunary-ai/lunary/pull/1060
• fix: tokenizers model order by @hughcrt in https://github.com/lunary-ai/lunary/pull/1062
• chore: bump langchain by @hughcrt in https://github.com/lunary-ai/lunary/pull/1063
• chore: add funding button by @hughcrt in https://github.com/lunary-ai/lunary/pull/1065
• feat: add postgres ssh tunneling in dev by @hughcrt in https://github.com/lunary-ai/lunary/pull/1066
• feat: integrate Vercel analytics and speed insights by @hughcrt in https://github.com/lunary-ai/lunary/pull/1068
• feat(python-sdk): better logging by @hughcrt in https://github.com/lunary-ai/lunary/pull/1069
• feat: better ingestion logging by @hughcrt in https://github.com/lunary-ai/lunary/pull/1070
• feat: update models by @hughcrt in https://github.com/lunary-ai/lunary/pull/1074
• feat(python-sdk): update consumer by @hughcrt in https://github.com/lunary-ai/lunary/pull/1073
• feat: better tests by @hughcrt in https://github.com/lunary-ai/lunary/pull/1076
• feat: org level api key by @hughcrt in https://github.com/lunary-ai/lunary/pull/1075
• chore: more descriptive auth errors by @hughcrt in https://github.com/lunary-ai/lunary/pull/1078
• fix: images build by @hughcrt in https://github.com/lunary-ai/lunary/pull/1079
• feat: org level analytics by @hughcrt in https://github.com/lunary-ai/lunary/pull/1077
• feat: clearable project dropdown in /team by @hughcrt in https://github.com/lunary-ai/lunary/pull/1080
• feat: improve org level analytics by @hughcrt in https://github.com/lunary-ai/lunary/pull/1081

​

What’s Changed

• feat: improve chart poper formating by @hughcrt in https://github.com/lunary-ai/lunary/pull/1052
• fix: tools serialization by @hughcrt in https://github.com/lunary-ai/lunary/pull/1053
• fix: google auth by @hughcrt in https://github.com/lunary-ai/lunary/pull/1056
• feat: better google auth error message by @hughcrt in https://github.com/lunary-ai/lunary/pull/1057

​

What’s Changed

• fix: message overflow by @hughcrt in https://github.com/lunary-ai/lunary/pull/1044
• fix: export message order by @hughcrt in https://github.com/lunary-ai/lunary/pull/1045
• feat: add gpt 5 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1050
• fix: traces related runs by @hughcrt in https://github.com/lunary-ai/lunary/pull/1051

​

What’s Changed

• feat: add opus 4.1 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1043

​

What’s Changed

• fix: export messages order by @hughcrt in https://github.com/lunary-ai/lunary/pull/1039
• feat: better cost badge in traces by @hughcrt in https://github.com/lunary-ai/lunary/pull/1040
• chore(deps): bump koa by @hughcrt in https://github.com/lunary-ai/lunary/pull/1041
• ci: fix submodule ref by @hughcrt in https://github.com/lunary-ai/lunary/pull/1042

​

What’s Changed

• feat: add metadata for conversations and messages by @hughcrt in https://github.com/lunary-ai/lunary/pull/1034
• feat: improve langchain ignore rules by @hughcrt in https://github.com/lunary-ai/lunary/pull/1035
• feat: improve model config modal by @hughcrt in https://github.com/lunary-ai/lunary/pull/1036

​

What’s Changed

• feat: add metadatas for retrievers by @hughcrt in https://github.com/lunary-ai/lunary/pull/1031
• feat: add barchart in custom charts by @hughcrt in https://github.com/lunary-ai/lunary/pull/1032

​

What’s Changed

• feat: otel pydantic ai by @hughcrt in https://github.com/lunary-ai/lunary/pull/1030

​

What’s Changed

• feat: add copy button to all messages by @hughcrt in https://github.com/lunary-ai/lunary/pull/1026
• feat: improve external users queries perf by @hughcrt in https://github.com/lunary-ai/lunary/pull/1028
• feat: add prompt testing against custom endpoint by @hughcrt in https://github.com/lunary-ai/lunary/pull/1027

​

What’s Changed

• feat: fix jwt expiration signout
• feat: add prompt testing against custom endpoint

​

What’s Changed

• chore(deps): bump urllib3 to 2.5.0 by @hughcrt in https://github.com/lunary-ai/lunary/pull/1012
• feat(python-sdk)/improve-ignore-list by @hughcrt in https://github.com/lunary-ai/lunary/pull/1010
• feat(python-sdk): save name for ignored elements by @hughcrt in https://github.com/lunary-ai/lunary/pull/1014
• feat: remove llm classifier by @hughcrt in https://github.com/lunary-ai/lunary/pull/1015
• feat: auto redirect on saml on join by @hughcrt in https://github.com/lunary-ai/lunary/pull/1016

​

What’s Changed

• feat: add missing backend zod validation
• feat: add Claude 4 pricing
• feat: add checklist api reference
• ci: fix helm publishing
• feat: tweak checklist endpoint
• doc: add auth to checklist api reference
• fix: provider config db constraint
• chore(deps-dev): bump h11 from 0.14.0 to 0.16.0 in /packages/python-sdk in the pip group across 1 directory by @dependabot in https://github.com/lunary-ai/lunary/pull/880
• chore: bump setuptools
• chore: remove custom script env variable
• chore: add security policy file
• ci: set gh action permission
• chore: do not capture 401 errors
• feat: add retry for end events in ingestion
• feat: sanitize sql input for alerts
• fix(py-sdk): custom project id conversations
• feat(py-sdk): improve ignore run
• chore(deps): bump requests from 2.32.2 to 2.32.4 in /packages/python-sdk in the pip group across 1 directory by @dependabot in https://github.com/lunary-ai/lunary/pull/1006
• feat: add o3 pro pricing

​

What’s Changed

• fix: xss vulnerability
• doc: audit logs api reference
• fix: exports

​

What’s Changed

• chore: add agents instructions by @hughcrt in https://github.com/lunary-ai/lunary/pull/981
• fix: IDOR patch by @hughcrt in https://github.com/lunary-ai/lunary/pull/982

​

What’s Changed

• feat: add sentry on frontend by @hughcrt in https://github.com/lunary-ai/lunary/pull/963
• feat: add tokenizer package by @hughcrt in https://github.com/lunary-ai/lunary/pull/965
• fix: delete conversation by @hughcrt in https://github.com/lunary-ai/lunary/pull/967
• codex: add bulk delete for llm logs by @hughcrt in https://github.com/lunary-ai/lunary/pull/970
• ci: build frontend only on releases by @hughcrt in https://github.com/lunary-ai/lunary/pull/975
• fix: google client id frontend self hosted by @hughcrt in https://github.com/lunary-ai/lunary/pull/977
• ci: fix ops submodule by @hughcrt in https://github.com/lunary-ai/lunary/pull/978
• ci: fix frontend build by @hughcrt in https://github.com/lunary-ai/lunary/pull/979

​

What’s Changed

• fix: prompt version update by @hughcrt in https://github.com/lunary-ai/lunary/pull/882
• feat: add batch run to dataset by @hughcrt in https://github.com/lunary-ai/lunary/pull/884
• ci: fix build by @hughcrt in https://github.com/lunary-ai/lunary/pull/885
• feat: improve row selection by @hughcrt in https://github.com/lunary-ai/lunary/pull/886
• chore: bump dependencies by @hughcrt in https://github.com/lunary-ai/lunary/pull/887
• feat: add cache for analytics query by @hughcrt in https://github.com/lunary-ai/lunary/pull/888
• chore: add turbopack in dev env by @hughcrt in https://github.com/lunary-ai/lunary/pull/889
• feat: make first load faster by @hughcrt in https://github.com/lunary-ai/lunary/pull/890
• feat: add more enrichers by @hughcrt in https://github.com/lunary-ai/lunary/pull/892
• feat: replace crisp by intercom by @hughcrt in https://github.com/lunary-ai/lunary/pull/894
• feat: improve analytics queries performance by @hughcrt in https://github.com/lunary-ai/lunary/pull/896
• feat: hide private key by default in UI by @hughcrt in https://github.com/lunary-ai/lunary/pull/897
• feat: add db index by @hughcrt in https://github.com/lunary-ai/lunary/pull/898
• feat: restructure sidebar by @hughcrt in https://github.com/lunary-ai/lunary/pull/899
• feat: add alert page by @hughcrt in https://github.com/lunary-ai/lunary/pull/900
• feat: make charts resizeable by @hughcrt in https://github.com/lunary-ai/lunary/pull/901
• feat: improve alerts by @hughcrt in https://github.com/lunary-ai/lunary/pull/903
• chore: add better sql debugging function by @hughcrt in https://github.com/lunary-ai/lunary/pull/904
• feat: analytics queries perfs greatly improved by @hughcrt in https://github.com/lunary-ai/lunary/pull/905
• feat: patch postgres library by @hughcrt in https://github.com/lunary-ai/lunary/pull/906
• feat: improve dashboard ux by @hughcrt in https://github.com/lunary-ai/lunary/pull/907
• feat: make top item clickable by @hughcrt in https://github.com/lunary-ai/lunary/pull/908
• feat: add topic filter by @hughcrt in https://github.com/lunary-ai/lunary/pull/909
• feat: improve playground by @hughcrt in https://github.com/lunary-ai/lunary/pull/910
• feat: add more dashboard filters by @hughcrt in https://github.com/lunary-ai/lunary/pull/912
• fix: openapi spec by @hughcrt in https://github.com/lunary-ai/lunary/pull/914
• feat: add chart level filters by @hughcrt in https://github.com/lunary-ai/lunary/pull/915
• feat: improve evaluators ux by @hughcrt in https://github.com/lunary-ai/lunary/pull/913
• feat: reorganize settings by @hughcrt in https://github.com/lunary-ai/lunary/pull/916
• feat: more evals by @hughcrt in https://github.com/lunary-ai/lunary/pull/918
• fix: evaluators by @hughcrt in https://github.com/lunary-ai/lunary/pull/921
• feat: improve evaluators by @hughcrt in https://github.com/lunary-ai/lunary/pull/922
• feat: switch toxicity eval to ml by @hughcrt in https://github.com/lunary-ai/lunary/pull/923
• feat: add prompt experiments by @hughcrt in https://github.com/lunary-ai/lunary/pull/924
• feat: allow edit chart name and description by @hughcrt in https://github.com/lunary-ai/lunary/pull/926
• feat: replace enrichments by evaluators by @hughcrt in https://github.com/lunary-ai/lunary/pull/927
• feat: add bar charts by @hughcrt in https://github.com/lunary-ai/lunary/pull/928
• feat: improve experiments ux by @hughcrt in https://github.com/lunary-ai/lunary/pull/929
• fix: client side error by @hughcrt in https://github.com/lunary-ai/lunary/pull/930
• fix: legacy evaluation endpoint by @hughcrt in https://github.com/lunary-ai/lunary/pull/931
• fix: signup by @hughcrt in https://github.com/lunary-ai/lunary/pull/932
• chore: clean logs by @hughcrt in https://github.com/lunary-ai/lunary/pull/933
• feat: tweak emails by @hughcrt in https://github.com/lunary-ai/lunary/pull/934
• ci: put back tests by @hughcrt in https://github.com/lunary-ai/lunary/pull/935
• feat: add smtp server connection check by @hughcrt in https://github.com/lunary-ai/lunary/pull/936
• ci: faster frontend builds by @hughcrt in https://github.com/lunary-ai/lunary/pull/937
• feat: less verbose invalid jwt logs by @hughcrt in https://github.com/lunary-ai/lunary/pull/939
• ci: cache playwright by @hughcrt in https://github.com/lunary-ai/lunary/pull/938
• ci: faster vercel builds by @hughcrt in https://github.com/lunary-ai/lunary/pull/940
• feat: experiments ux improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/941
• fix: legacy evaluation and checklists route by @hughcrt in https://github.com/lunary-ai/lunary/pull/942
• feat: move data retention to data rules by @hughcrt in https://github.com/lunary-ai/lunary/pull/943
• fix: old export by @hughcrt in https://github.com/lunary-ai/lunary/pull/944
• feat: experiments ux improvements by @hughcrt in https://github.com/lunary-ai/lunary/pull/945
• feat: add beta in sidebar for experiments by @hughcrt in https://github.com/lunary-ai/lunary/pull/946
• fix: enrichers by @hughcrt in https://github.com/lunary-ai/lunary/pull/947
• fix: experiments layout by @hughcrt in https://github.com/lunary-ai/lunary/pull/948
• fix: pause subscriptions upgrade by @hughcrt in https://github.com/lunary-ai/lunary/pull/949
• fix: stripe webhooks by @hughcrt in https://github.com/lunary-ai/lunary/pull/950
• fix: subscription update by @hughcrt in https://github.com/lunary-ai/lunary/pull/951
• fix: ingestion by @hughcrt in https://github.com/lunary-ai/lunary/pull/957
• feat: add mistral in get started examples by @hughcrt in https://github.com/lunary-ai/lunary/pull/958
• feat: add sentry on backend by @hughcrt in https://github.com/lunary-ai/lunary/pull/960
• fix: dashboard charts spans by @hughcrt in https://github.com/lunary-ai/lunary/pull/961
• feat: allow to change chart type by @hughcrt in https://github.com/lunary-ai/lunary/pull/962

​

What’s Changed

​

What’s Changed

• feat: improve trace query perfs
• fix: filters
• fix: trace query
• fix: conversations
• fix: neighbor query
• fix: google login

​

What’s Changed

• fix: migrations
• feat: add secondary click on traces and conversations
• feat: stronger email verification policies
• feat: improve auth join flow
• feat: add sso on join page

​

What’s Changed

• feat(python-sdk): add py.typed file
• feat: faster metadata filters
• feat: better metadata filter value parsing
• feat: db queries performance improvements
• feat: add new openai models
• feat: add traces in conversations exports
• feat: add cost refresh button
• feat: improve settings page performance
• feat: improve settings ux

​

What’s Changed

• fix: trace overflow
• feat: add gpt 4.1 pricing
• feat: add gpt 4.5 models to playground

​

What’s Changed

• feat: add empty feedback filter
• feat: add total feedback charts

​

What’s Changed

• feat(python sdk): add ssl verify option by @hughcrt in https://github.com/lunary-ai/lunary/pull/839
• feat: add latest gemini models by @hughcrt in https://github.com/lunary-ai/lunary/pull/843
• ci: improve build speed by @hughcrt in https://github.com/lunary-ai/lunary/pull/844
• Fixed full documentation URL by @KennyStryker in https://github.com/lunary-ai/lunary/pull/840
• chore(deps-dev): bump langchain-core from 0.3.13 to 0.3.15 in /packages/python-sdk in the pip group across 1 directory by @dependabot in https://github.com/lunary-ai/lunary/pull/842
• feat: add support for openai prompt caching by @hughcrt in https://github.com/lunary-ai/lunary/pull/845
• ci: rename build step by @hughcrt in https://github.com/lunary-ai/lunary/pull/846
• ci: fix build by @hughcrt in https://github.com/lunary-ai/lunary/pull/847

​

New Contributors

• @KennyStryker made their first contribution in https://github.com/lunary-ai/lunary/pull/840

​

What’s Changed

• fix: enrichments page rendering

​

What’s Changed

• feat: add whitelisted domain recaptcha
• feat: allow export with private key
• doc: run export
• fix: conversations feedback filters

​

What’s Changed

• feat: improve ingestion rules

​

What’s Changed

• fix: docker compose autoheal

​

What’s Changed

• feat: clean settings
• chore: update documentation
• feat: allow prompt creation without variation for datasets
• feat: improve dataset endpoints
• feat: js language detection
• feat: security history
• feat: improve data retention
• feat: improve enrichments
• fix: sql migration 0064

​

What’s Changed

• chore: remove npm from docker images

​

What’s Changed

• fix: recaptcha for self hosted version

​

What’s Changed

• feat: add claude 3 7 sonnet
• fix: add prompt to dataset
• feat: add custom provider
• feat: remove save api key button
• chore: upgrade bun
• feat: improve datasets api
• doc: update score definition openai
• feat: return full dataset api
• fix: openapi definitions
• feat: better openapi schema
• feat: better user facing errors
• feat: add bedrock to custom providers
• feat: update bun
• fix: anthropic prompt playground

​

What’s Changed

• fix: charts custom events by @hughcrt in https://github.com/lunary-ai/lunary/pull/802
• feat: add custom chart description by @hughcrt in https://github.com/lunary-ai/lunary/pull/803

​

What’s Changed

• fix: dataset prompt creation by @hughcrt in https://github.com/lunary-ai/lunary/pull/789
• chore(python-sdk): update repo link by @hughcrt in https://github.com/lunary-ai/lunary/pull/791
• doc(python-sdk): update homepage by @hughcrt in https://github.com/lunary-ai/lunary/pull/792
• feat: add custom dashoards for everybody by @hughcrt in https://github.com/lunary-ai/lunary/pull/795
• fix: user count in charts by @hughcrt in https://github.com/lunary-ai/lunary/pull/794
• fix: pipeline by @hughcrt in https://github.com/lunary-ai/lunary/pull/799
• fix: page switching from prompts by @hughcrt in https://github.com/lunary-ai/lunary/pull/800

​

What’s Changed

• feat: deepseek v3

​

What’s Changed

• feat: better perf agent chart by @hughcrt in https://github.com/lunary-ai/lunary/pull/780
• fix: backend image by @hughcrt in https://github.com/lunary-ai/lunary/pull/781
• fix: submodule sync by @hughcrt in https://github.com/lunary-ai/lunary/pull/784
• feat: add signup captcha by @hughcrt in https://github.com/lunary-ai/lunary/pull/785
• feat: allow message copy by @hughcrt in https://github.com/lunary-ai/lunary/pull/786
• fix: crash charts by @hughcrt in https://github.com/lunary-ai/lunary/pull/787

​

What’s Changed

• feat: improve feedback filters perfs by @hughcrt in https://github.com/lunary-ai/lunary/pull/770
• feat: improve model list query perf by @hughcrt in https://github.com/lunary-ai/lunary/pull/771
• feat: improve feedback filer by @hughcrt in https://github.com/lunary-ai/lunary/pull/773
• fix: migration by @hughcrt in https://github.com/lunary-ai/lunary/pull/772
• fix(python-sdk): reconciliation in async context by @hughcrt in https://github.com/lunary-ai/lunary/pull/774

​

What’s Changed

• fix: load more by @hughcrt in https://github.com/lunary-ai/lunary/pull/769

​

What’s Changed

• feat: over 30x performance improvement for loading the LLM calls table @hughcrt in https://github.com/lunary-ai/lunary/pull/768

​

What’s Changed

• update demo video by @vincelwt in https://github.com/lunary-ai/lunary/pull/762
• update readme by @vincelwt in https://github.com/lunary-ai/lunary/pull/763
• feat: add github link in sidebar by @hughcrt in https://github.com/lunary-ai/lunary/pull/764
• feat: add tracking notification dismissed by @hughcrt in https://github.com/lunary-ai/lunary/pull/765
• feat: improve parent feedback query by @hughcrt in https://github.com/lunary-ai/lunary/pull/766
• feat: drop parent feedback cache materialized view by @hughcrt in https://github.com/lunary-ai/lunary/pull/767

​

What’s Changed

• fix: patch postgres package by @hughcrt in https://github.com/lunary-ai/lunary/pull/753
• fix: column icon overlap by @hughcrt in https://github.com/lunary-ai/lunary/pull/754
• fix: improve enrichers by @hughcrt in https://github.com/lunary-ai/lunary/pull/755
• feat: remove enrichers from sidebar by @hughcrt in https://github.com/lunary-ai/lunary/pull/756
• feat: user names are now clickable by @hughcrt in https://github.com/lunary-ai/lunary/pull/757
• feat: improve chevron styling by @hughcrt in https://github.com/lunary-ai/lunary/pull/758
• feat: add developer role by @hughcrt in https://github.com/lunary-ai/lunary/pull/759
• fix: styling column icon by @hughcrt in https://github.com/lunary-ai/lunary/pull/760
• fix: feedback in conversations by @hughcrt in https://github.com/lunary-ai/lunary/pull/761

​

What’s Changed

• feat: add top agents chart by @hughcrt in https://github.com/lunary-ai/lunary/pull/738
• fix: user list crash by @hughcrt in https://github.com/lunary-ai/lunary/pull/739
• feat: better formatting chart tooltip by @hughcrt in https://github.com/lunary-ai/lunary/pull/740
• fix: chart selector aggregation by @hughcrt in https://github.com/lunary-ai/lunary/pull/741
• fix: chart ticks by @hughcrt in https://github.com/lunary-ai/lunary/pull/742
• feat: display trace score in datatable by @hughcrt in https://github.com/lunary-ai/lunary/pull/743
• fix: rows ordering by @hughcrt in https://github.com/lunary-ai/lunary/pull/744
• fix: tooltips by @hughcrt in https://github.com/lunary-ai/lunary/pull/745
• feat: better formatting tooltip on hourly granularity by @hughcrt in https://github.com/lunary-ai/lunary/pull/746
• chore: switch to bun by @hughcrt in https://github.com/lunary-ai/lunary/pull/747
• feat: add o1 pricing by @hughcrt in https://github.com/lunary-ai/lunary/pull/748
• fix: frontend build by @hughcrt in https://github.com/lunary-ai/lunary/pull/750
• fix: enrichers query by @hughcrt in https://github.com/lunary-ai/lunary/pull/749
• ci: enrichers build by @hughcrt in https://github.com/lunary-ai/lunary/pull/751

​

What’s Changed

• feat: improve llm calls list performances by @hughcrt in https://github.com/lunary-ai/lunary/pull/736
• feat: conversations chart by @hughcrt in https://github.com/lunary-ai/lunary/pull/737

​

What’s Changed

• fix: overflow user with long names
• feat: add top topics chart
• feat: add topics filter

​

What’s Changed

• fix: scores
• feat: add pricing for openrouter anthropic
• fix: cached llm cost
• feat: make user dropdown data loaded dynamically

​

What’s Changed

• feat: add perplexity models to playground
• Add sonar-reasoning
• feat: add o3 to playground
• fix: project switching bug

​

What’s Changed

• fix: metadata keys query by @hughcrt in https://github.com/lunary-ai/lunary/pull/724

​

What’s Changed

• feat: add phi4 and deepseek in model list
• feat: add ibm watsonx.ai models
• feat: add background index creation job system

​

What’s Changed

• feat(py-sdk): add lc error placeholder
• ci: change tag name for cloud releases
• feat: add more llama models
• feat: add more models in prompts
• feat: put back llama 3.1-8b
• feat: add perplexity model

​

What’s Changed

• feat: put back checklists
• feat(py-sdk): pass custom model name to langchain
• feat: go to next previous run or trace

​

What’s Changed

• feat: redirect on dashboard delete
• fix(python-sdk): anthropic example
• chore(python-sdk): update dependencies
• fix: analytics role

​

What’s Changed

• fix: filters
• chore: remove backend version in ui for cloud version
• feat: hide parent run id in metadata ui
• fix: user charts
• feat: add ibm models pricing
• chore: add python sdk in monorepo
• feat(python-sdk): add integration with IBM wastonx.ai foundation models

​

What’s Changed

• feat: remove evaluation playground

​

What’s Changed

• fix: full textsearch performances

​

What’s Changed

• feat: better tests
• feat: add env variable to disable feedback cache
• ci: improve test pipeline
• fix: security patches

​

What’s Changed

• fix: chat ingestion foreign key error

​

What’s Changed

• chore: remove console.log

​

What’s Changed

• ci: do no deploy on release
• ci: better deployment script

​

What’s Changed

• fix: patch package postgres

​

What’s Changed

• fix: padding auth page
• fix: put back ci tests
• ci: improve ci tests workflow speed

​

What’s Changed

• fix: trace error

​

What’s Changed

• fix: tiktoken error on ingestion
• fix: chat export

​

What’s Changed

• fix: redirection bug

​

What’s Changed

• feat: audio input
• feat: custom dashboards

​

What’s Changed

• feat: better onboarding
• fix: clickjacking
• fix: views

​

What’s Changed

• test: logs csv export
• feat: add Github OAuth
• fix: ingestion
• ci: fix helm release tagging

​

What’s Changed

• chore(deps): bump cross-spawn from 7.0.3 to 7.0.5 in the npm_and_yarn group across 1 directory
• fix: log export limit

​

What’s Changed

• fix: view checks deserialization
• fix: run query

​

What’s Changed

• feat: improve export streaming
• fix: remove horizontal scrollbar in dropdown
• fix: default date picker value
• feat: better access control granularity

​

What’s Changed

• feat: users avatar in threads
• fix: llm call json rendering
• fix: ingestion

​

What’s Changed

• fix: traces rendering
• fix: security patches

​

What’s Changed

• fix: readme
• fix: security patch

​

What’s Changed

• fix: topic enricher
• fix: thread rendering
• fix: scores

​

What’s Changed

• feat: color harmony
• feat: add streaming for export
• feat: add custom scoring on runs
• feat: add custom event tracking in threads

​

What’s Changed

• feat: google login
• fix: signup tracking
• fix: session and local storage clearing on signup
• fix: security patch

​

What’s Changed

• feat: improve user filter
• fix: multi select do not close

​

What’s Changed

• ci: restrict auto review PRs
• fix: security patches

​

What’s Changed

• fix: cost for character based mappings
• fix: sentry make signup redirect crash
• feat: implement streaming from database for csvjsonl exports
• fix: flowise legacy tracking
• fix: update view columns
• feat: update gpt4o model in playground

​

What’s Changed

• ci: add ml image build
• ci: add helm publishing
• ci: update legacy dockerfile env syntax
• fix: traces cost

​

What’s Changed

• feat: nested multi-level trace objects and export support in threads
• chore: dockerfile legacy key value
• feat: allow admin to delete projects

​

What’s Changed

• chore: add missing types for swagger
• fix: all OpenAPI paths
• chore: shorten some api routes doc title
• fix: enrichers service
• fix: enrichments columns dupplicates on project switching
• feat: add collaborator role
• feat: hide create template button for unauthorized roles
• ci: add auto pr workflow
• fix: smart data filtering UI + ingestion

​

What’s Changed

• feat: db delete child run cascade
• feat: add delete run, thread and trace button

​

What’s Changed

• feat: improve design of auth pages
• fix: auth dark mode
• feat: OpenAPI spec
• fix: backend sentry version
• chore: downgrade node version
• feat: display current version in UI
• fix: lunary frontend version displaying on Vercel
• fix: automatically set saml_enabled

​

What’s Changed

• feat: session expired redirection

​

What’s Changed

• word wrap and tooltip for long project names
• fix: checklists patch route body parsing
• chore(deps-dev): bump rollup from 4.22.0 to 4.22.4 in the npm_and_yarn group across 1 directory
• fix: analytics viewer role

​

What’s Changed

• fix: openai to anthropic tool conversion

​

What’s Changed

• fix: security patch
• feat: add configurable rate limiting

​

What’s Changed

• fix: integer too large error
• chore: improve docker build pipeline
• chore: update readme
• feat: better sentiment analysis

​

What’s Changed

• fix: anthropic tool nested objects

​

What’s Changed

• fix: anthropic tool calls playground

​

What’s Changed

• fix: redos
• fix: data warehouse connector access control
• fix: checklist access control

​

What’s Changed

• feat: add gpt-4o-2024-05-13 to playground

​

What’s Changed

• chore: improve enrichers speed
• fix: playground anthropic tools

​

What’s Changed

• feat: add OpenAI new pricing
• fix: calc cost
• fix: redos

​

What’s Changed

• feat: allow a team owner to make another user owner
• fix: query inconsistency causing routing bug

​

What’s Changed

• fix: prompt playground claude

​

What’s Changed

• feat: add data warehouse settings
• fix: migrations

​

What’s Changed

• chore: add prettier semicolon
• fix: selected columns resetting on user selection
• fix: top Models cards not showing empty data placeholder
• fix: analytics filters for tops
• feat: add enrichments page

​

What’s Changed

• fix: playground tools

​

What’s Changed

• fix: email signature

​

What’s Changed

• fix: metadata persistence
• feat: put back edit checklist
• fix: sorting by token
• fix: edit checklist
• feat: adapt emails for selfhosting

​

What’s Changed

• feat: remove radar page
• feat: live prompts use publishedAt date rather than the createdAt date
• fix: email sanitization
• feat: more flexible cors policy
• feat: better error message user invitation

​

What’s Changed

• feat: increase payload limit to 20mb

​

What’s Changed

• feat: remove payload body limit

​

What’s Changed

• feat: switch from resend to SMTP client
• fix: saml acs

​

What’s Changed

• fix: tag dropdown
• fix: reset password
• feat: add prompts by default for selfhosting

​

What’s Changed

• fix: enrichment render loop
• fix: badge rendering
• fix: gmail email parsing
• fix: cors too permissive
• fix: add org level permissions
• fix: related run security patch
• feat: name sanitization in emails

​

What’s Changed

• fix: SAML security patch
• content filtering
• fix: add missing access control
• fix: password reset security patch
• fix: SSRF security patch
• fix: XSS security patch
• fix: accept anthropic tool calling format
• fix: user invite security patch
• fix: tool list ui in traces

​

What’s Changed

• fix: owner_id not nullable
• fix: gpt-4o-mini pricing
• fix: add missing free plan paywall
• fix: render enrichment
• fix: dataset permissions check

​

What’s Changed

• feat: add analytics role by @hughcrt in https://github.com/lunary-ai/lunary/pull/491

​

What’s Changed

• fix: user column sorting
• fix: evaluation result error
• feat: add data retention policy
• feat: improve search
• feat: add gpt-4o-2024-08-06 to costs mapping
• chore: remove teamseats sentry error
• feat: license improvements

​

What’s Changed

• fix: text template deployment

​

What’s Changed

• feat: make long traces scrollable
• fix: assistant message displaying
• feat: add open parent trace button
• fix: sanitize runID before countTokens
• fix: page crashing for uncompleted runs
• fix: metadata ingestion

​

What’s Changed

• ci: fix ml add helm workflow by @hughcrt in https://github.com/lunary-ai/lunary/pull/472

​

What’s Changed

• ci: add ml image in release by @hughcrt in https://github.com/lunary-ai/lunary/pull/470
• ci: fix ml and add helm auto publish by @hughcrt in https://github.com/lunary-ai/lunary/pull/471

​

What’s Changed

• fix: query param vulnerability
• fix: vulnerabilities
• ci: add ml image in release

​

What’s Changed

• perf: model names
• perf: tag list
• perf: metadata key list
• refactor: remove filter feedback
• refactor: remove unused materialized view
• fix: cal run cost by

​

What’s Changed

• fix: ingestion for multiple langchain callbacks by @hughcrt in https://github.com/lunary-ai/lunary/pull/451
• fix: misc errors by @hughcrt in https://github.com/lunary-ai/lunary/pull/452
• fix: ingestion by @hughcrt in https://github.com/lunary-ai/lunary/pull/455
• feat: add timeout for tokenizer by @hughcrt in https://github.com/lunary-ai/lunary/pull/456
• feat: improve token counting by @hughcrt in https://github.com/lunary-ai/lunary/pull/457
• perf: remove unused postgres indexes by @hughcrt in https://github.com/lunary-ai/lunary/pull/458
• perf: run and external_users queries by @hughcrt in https://github.com/lunary-ai/lunary/pull/459

​

What’s Changed

• test: fix failling tests
• fix: calc cost
• fix: gh workflow
• fix: spelling comment
• fix: realtime evaluators
• feat: column sorting
• fix: migrations
• feat: add gpt-4o-mini support

​

What’s Changed

• fix: docker build

​

What’s Changed

• fix: build

​

What’s Changed

• fix: build

​

What’s Changed

• feat: eval performances

​

What’s Changed

• feat: remove custom models paywall

​

What’s Changed

• chore: remove auto pr
• fix: docker frontend NEXT_PUBLIC env variable

​

What’s Changed

• feat: revamp UX
• fix: user list overflow
• fix: not ingested ui
• fix: settings rendering
• fix threads
• feat: update logs filters
• fix: run count query
• fix: threads rendering
• feat: add metadata ingestion for end events
• fix: date filter
• feat: rename templateId to templateVersionId
• feat: add dynamic seat allowance
• fix: docker frontend build
• fix: views
• fix: smart data filtering access

​

What’s Changed

• Fix: stripe canceled webhook
• fix: ingestion
• fix: data filtering

​

What’s Changed

• feat: add support for for sonnet 3.5
• fix: unclickable menu items
• feat: improve analytics
• Feat: evaluation results improvements
• fix: invalid CSV export
• fix/stripe update webhook
• feat: license config for smart data filtering for self hosted version

​

What’s Changed

• feat: get latest template endpoint

​

What’s Changed

• fix: lunary-ee by @hughcrt in https://github.com/lunary-ai/lunary/pull/383
• perf: optimize runs query by @hughcrt in https://github.com/lunary-ai/lunary/pull/385
• perf: text search for runs by @hughcrt in https://github.com/lunary-ai/lunary/pull/387
• style: runcate long model names by @7HR4IZ3 in https://github.com/lunary-ai/lunary/pull/388
• feat: add modal for editing long variable values by @7HR4IZ3 in https://github.com/lunary-ai/lunary/pull/380

​

New Contributors

• @7HR4IZ3 made their first contribution in https://github.com/lunary-ai/lunary/pull/388

​

What’s Changed

• Advanced analytics released 🎉

​

What’s Changed

• feat: update lunary-ee start script
• fix: large traces UI
• Feat: JSONL fine-tune exports
• fix: radar

​

What’s Changed

• fix: migrations

​

What’s Changed

• feat: analytics next
• feat: button to remove user data
• fix: agent summary

​

What’s Changed

• fix: stripe error self hosted

​

What’s Changed

• fix: self hosting images

​

What’s Changed

• fix: lunary-ee

​

What’s Changed

• fix: make run private
• fix: pii radar
• fix: lunary-ee
• fix stripe webhook

​

What’s Changed

• fix: email spam
• fix: text template deployment
• fix: playground api route default variables
• ci: put back docker cloud build
• fix: user invitation

​

What’s Changed

• fix: radar
• feat: improve Google SAML support
• fix: postgres uncamel
• fix: backend-docker-build
• feat: prompt notepad
• feat: open in playground copy tools & output
• fix: SQL events limit [LLM-1385]
• fix: project access

​

What’s Changed

• ci: put back ml build
• fix copy/pasting SAML
• replace cal.com by savvycal
• fix: radar empty run id
• fix: backend plan check
• fix: api security by @hughcrt in https://github.com/lunary-ai/lunary/pull/330

​

What’s Changed

• fix: evaluations
• fix: user display
• ci: publish arm docker images

​

What’s Changed

• feat: allow to update a run in ingestion
• ci: improvement
• ci: improvement
• feat: team plan
• feat: prompt improvements
• fix: template version
• fix: filters url deserialization
• ci: improve deployment speed
• feat: add gpt4o pricing
• feat: add gpt4o in prompt playground
• chore: update posthog

​

What’s Changed

• fix: radars
• feat: enable ssh for azure docker
• feat: cached call ui
• fix: lunary-ee image
• feat: add tags for all run types in trace
• doc: update env example

​

What’s Changed

• doc: update local running instructions
• ci: fix version publish
• fix: feedback radar
• perf: radars

​

What’s Changed

• ci: remove arm64
• fix: filters return data of wrong type

​

What’s Changed

• fix: skip email verify for self hosting
• feat: allow lunary ee to use one port

​

What’s Changed

• fix: prompt deletion
• feat: put back basic
• fix: skip email verify for self hosting

​

What’s Changed

• fix: lunary-ee docker build
• fix: azure openai stream on backend

​

What’s Changed

• fix: feedback filter

​

What’s Changed

• feat: ui improvements
• perf: database and api
• fix: prompt playground with Azure OpenAI

​

What’s Changed

• feat: add feedback from ui
• feat: add where did you find us in signup
• fix: team full when not actually full
• fix: project select for not allowed org
• fix: user allowance
• ci: improve deployment speed
• feat: increase request body size limit
• fix: feedback

​

What’s Changed

• fix: stray a
• style: improve check filters display
• style: improve check filters display
• fix: and or dropdown render

​

What’s Changed

• fix: load run from url even if not in project or not loaded yet
• build: add arm64 to docker images builds
• fix: user filter

​

What’s Changed

• fix: external user list
• feat: add azure openai support

​

What’s Changed

• fix: delete dataset
• feat: new OpenAI models
• fix: ml deployment
• ci: auto pr creation
• ci: do not trigger pr on new tag
• fix: access public run
• chore: remove trailing ? in url for /logs
• style: add loading spinner on log pane
• fix: decimal number deserialization in checks
• fix: analytics user usage

​

What’s Changed

• fix: pipeline
• fix: ml deployment
• feat: license key check
• chore: prepare release
• chore: prepare release
• fix: 404 with new dataset

​

What’s Changed

• fix: column billing does not exist
• chore: update ops
• fix: check plan before changing user role
• feat: jwt invalidation
• fix: radars
• chore: update pipeline
• chore: build images in parallel
• fix: pipeline
• fix: rate limit password reset
• fix: authorization delete dataset

​

What’s Changed

• fix: set user to owner
• fix: backend plan check for radar and evaluation creation

​

What’s Changed

• stricter webhook
• feat: add sentry on frontend
• fix: stream prompt
• fix: prompt response stream
• fix: user payload

​

What’s Changed

• fix: metadata rendering error
• fix: infinite render

​

What’s Changed

• chore: auto build images on release
• chore: replace deprectad gh actions command
• fix: get evaluation result
• fix: modify owner

​

What’s Changed

• chore: docker version bump
• Fix: prevent admin to modify owner
• feat: radar timeout on error
• chore: remove namespace.co
• chore: update monolithic image
• Fix: Remove password hash from payload
• chore: cache playwright binary in github actions
• chore: auto build images on new release

​

What’s Changed

• feat: join improvements
• Feat: metadata ingestion & filter
• Fix: old prompt variables
• fix: github actions
• Feat: feedback cascading, filters v2, Users page search & pagination
• chore: pipeline improvements
• fix: invite link
• fix: request password reset
• fix: frontend build
• fix: pipeline
• Feat: private API keys
• fix: public routes
• fix: login when invalid token
• feat: add ml health check endpoint
• fix: tools calls issues
• Feat: better eval results (fixed branch)
• feat: multi arch docker

​

What’s Changed

• Feat: trace improvements
• Feat: Improve retrievers UI
• Feat: playwright tests by
• Feat: fix Claude 3 and mistral support
• Feat: prompt menu improvements

• Fix cost tracking with fine-tuned models
• Fix error in SQL setup
• Fix featurebase crashing self-hosted

• 🔍 Search & filters in dashboard
• 🗃️ Exports (on pro plans)
• 🔈 Feedback tracking in general availability
• 🐛A lot of bug fixes and stabilization across the board, esp. on the JS and Python SDKs, like counting tokens for streaming queries.
• 📜 Revamped documentation. It’s now much clearer how to setup user tracking tags, and feedback tracking.

• New sign up flow
• Webhook to automatically create first app on signup
• Infinite loading & performance improvements for agents & requests tables
• Better UI to view chat messages and errors
• Add endpoint to check app’s existence

• Fix page stuttering & other UI quirks
• Fix rendering certain chats items

• User tracking
• Cost monitoring
• Lots of design & UI improvements
• Lots of fixes in the dashboard

• Basic features
• Agent tracing
• Basic analytics
• Langchain Tracer API support