What's Changed
- Feat: Search Eval Testing Overhaul (provide ground truth, categorize query, etc.) by @Orbital-Web in #4739
- fix: Fix chats dragging issue by @raunakab in #4751
- improve impersonation logging slightly by @rkuo-danswer in #4758
- add k8s probes by @rkuo-danswer in #4752
- fix tool calling for bedrock claude models by @evan-onyx in #4761
- add probe signals for beat by @rkuo-danswer in #4760
- add probes by @rkuo-danswer in #4762
- Fix POSTGRES_IDLE_SESSIONS_TIMEOUT by @Weves in #4765
- v1 refresh drive creds during perm sync by @evan-onyx in #4768
- Fix/add back search with files by @Weves in #4767
- print sanitized api key to help troubleshoot by @rkuo-danswer in #4764
- refactor: stream_llm_answer by @Orbital-Web in #4772
- fix: Create new grouping for CRM connectors by @raunakab in #4776
- Feature/tenant reporting 2 by @rkuo-danswer in #4750
- Feature/helm k8s probes 2 by @rkuo-danswer in #4766
- Bugfix/slack bot debugging by @rkuo-danswer in #4782
- connector filter bug fix by @Subash-Mohan in #4771
- skip excel files that openpyxl fails on by @evan-onyx in #4787
- new slack rate limiting approach by @rkuo-danswer in #4779
- fix: Remove "Refining Answer" popup by @raunakab in #4783
- add a comment by @rkuo-danswer in #4789
- Fix frozen add token rate limit migration by @Weves in #4794
- Simplify passing in of file IDs for filtering by @Weves in #4791
- jira daylight savings handling by @evan-onyx in #4797
- Fix failed docs table by @Weves in #4800
- try fixing slack bot by @rkuo-danswer in #4792
- drive perm sync logs + misc deployment improvements by @evan-onyx in #4788
- Adjust user file access by @Weves in #4810
- Remove ordering-only flow by @Weves in #4812
- [Connectors][Script] Resume Paused Connectors by @trial-danswer in #4798
- Adjust script by @Weves in #4813
- Fix hard delete of agentic chats by @Weves in #4803
- fix: updating gemini models by @CaptainJeff in #4806
- bumping docker push action version by @rkuo-danswer in #4816
- Improve reasoning detection by @Weves in #4817
- New metadata for Jira for KG by @joachim-danswer in #4785
- Feature/vespa bump by @rkuo-danswer in #4819
- Small embedding model cleanups by @Weves in #4820
- Small tweak to user invite flow by @Weves in #4825
- Increase of Agent LLM Timeouts by @joachim-danswer in #4821
- Upgrade asyncpg for Python 3.12 by @maver1ck in #4699
- Update GitHub Connector metadata by @Orbital-Web in #4769
- add slack percentage progress by @rkuo-danswer in #4809
- Enable default quantization by @Weves in #4815
- Extend the onyx_vespa_schemas.py script by @Weves in #4835
- Fixed indexing when no sites are specified by @wenxi-onyx in #4822
- remove Hagen from CONTRIBUTING.md by @rexjohannes in #4778
- Restart script clarity by @wenxi-onyx in #4837
- Restart script clarity by @wenxi-onyx in #4839
- Knowledge Graph v1 by @joachim-danswer in #4626
- Kg batch clustering by @Orbital-Web in #4847
- KG autofill metadata by @Orbital-Web in #4834
- IAM support for S3 / blob connector by @Subash-Mohan in #4846
- Order sub-questions for chat message by @joachim-danswer in #4848
- fix: Query History Export by @raunakab in #4841
- fix(discourse): Remove early break that was limiting topics to indexing batch size by @suvodhoy in #4855
- Enable embedding parallelism by @Weves in #4850
- errors have correct file id by @evan-onyx in #4818
- drive smaller checkpoints v1 by @evan-onyx in #4849
- Add perm sync to indexing for google drive by @Weves in #4842
- Enhancement/gpt4o image gen support by @Subash-Mohan in #4859
- Add Qwen icon in LLM list and update provider icon mapping by @ZhipengHe in #4625
- remove lru cache by @evan-onyx in #4865
- fix: Add rate-limiting to Teams API request by @raunakab in #4854
- feat: default kg entity types by @Orbital-Web in #4851
- fix: Change how replies are processed by @raunakab in #4870
- feat: add configurable image model name and update dependencies by @Subash-Mohan in #4873
- fix: Remove threading from Teams connector by @raunakab in #4875
- Switch to chonkie from llamaindex chunker by @Weves in #4838
- fix: kg answer streaming by @Orbital-Web in #4877
- skip large empty excel files by @evan-onyx in #4878
- kg_update by @joachim-danswer in #4858
- Dual search pipeline for non-tool-calling LLMs by @joachim-danswer in #4872
- cleanup: Edit logic for default entity-types population by @raunakab in #4876
- bump disk size by @rkuo-danswer in #4882
- fix kg variable by @Orbital-Web in #4881
- Prep KG on-demand indexing through celery by @joachim-danswer in #4874
- new script for hard deleting sessions by @wenxi-onyx in #4883
- Make password validation configurable by @Django149 in #4869
- feat: Knowledge graph full-stack implementation by @raunakab in #4790
- Bugfix/honor disable default slack config by @wenxi-onyx in #4891
- Remove non-helm kubernetes deployment option by @Weves in #4904
- Kg Subtype Rework by @Orbital-Web in #4892
- fix db connection assertion by @evan-onyx in #4905
- Bugfix/salesforce correctness 3 by @rkuo-danswer in #4598
- feat: Add support for Assume Role authentication in S3 by @Subash-Mohan in #4907
- fix: Implement time-filtering for MS Teams document fetching by @raunakab in #4906
- KG - Entity-Only Path by @joachim-danswer in #4898
- fail loudly when error should be propagated by @evan-onyx in #4903
- tiny connector logging tweaks by @evan-onyx in #4908
- Feature/connector creation feedback by @trial-danswer in #4644
- Add psql to backend containers by @Weves in #4909
- ADd slackbot to helm by @Weves in #4911
- build: Remove ruff by @raunakab in #4912
- Kg config refactor by @Orbital-Web in #4902
- fix(discourse-connector): handle redirect issue by @suvodhoy in #4780
- fix celery beat from scheduling ee tasks by @Orbital-Web in #4915
- Really add psql to api-server by @Weves in #4919
- GCS metadata processing by @wenxi-onyx in #4879
- fix gemini model names + add vertex claude sonnet 4 by @wenxi-onyx in #4920
- S3 like file store by @Weves in #4897
- Fix no subject gmail docs by @wenxi-onyx in #4922
- KG Updates by @joachim-danswer in #4925
- Forcing vespa language by @joachim-danswer in #4926
- feat: comma separated citations by @Orbital-Web in #4923
- Hubspot connector enhancements by @Weves in #4927
- KG Vespa Rate Limit Fix by @Orbital-Web in #4931
- build: Update ruff to latest version by @raunakab in #4914
- fix: Add minor type-checking fixes by @raunakab in #4916
- Fix parallel Search tool calls by @Weves in #4935
- Fix query history by @Weves in #4937
- docker dev and prod template by @joachim-danswer in #4936
- fix: Fix sidebar logic to hit kg-endpoint for admins only by @raunakab in #4938
- Add support for db proxy by @Weves in #4932
- Fix single-tenant normalization by @joachim-danswer in #4942
- Handle very long file names by @Weves in #4939
- KG extraction without vespa queries by @joachim-danswer in #4940
- Add MinIO configuration to env template and update restart script for… by @Subash-Mohan in #4944
- Update Docker Compose restart policy to unless-stopped by @DAcodedBEAT in #4945
- curator bug fixes by @wenxi-onyx in #4941
- feat: JIRA doc syncing by @raunakab in #4899
- KG Attribute Overhaul + Processing Tests by @Orbital-Web in #4933
- Add creator to linear by @Orbital-Web in #4948
- Extend (fix) definition of public docs in KG Views by @joachim-danswer in #4950
- Fix query history 2 by @Weves in #4951
- kg cleanup + reintroducing deep extraction & classification by @Orbital-Web in #4949
- feat: MS Teams permission syncing by @raunakab in #4934
- perf: Improve performance of MS Teams permission-syncing logic by @raunakab in #4953
- Improve drive group sync by @Weves in #4952
- KG vespa error handling + separating relationship transfer & vespa updates by @Orbital-Web in #4954
- kg source fix by @Orbital-Web in #4956
- Fix test_gdrive_perm_sync_with_real_data patching by @Weves in #4961
- Improve mt migration process by @Weves in #4960
- fix: Fix bug in which emails would be fetched during initial indexing by @raunakab in #4959
- Remove CONCURRENTLY from migrations by @Weves in #4965
- Move onyx_list_tenants.py to make sure it's in the image by @Weves in #4966
- perm sync validation framework by @evan-onyx in #4958
- Reduce frequency of heavy checks on primary for cloud by @Weves in #4967
- Onyx Metadata Header for File Connector by @yuhongsun96 in #4968
- search speedup by @evan-onyx in #4971
- Small gdrive perm sync enhancement by @Weves in #4973
- account for category prefix added by user by @wenxi-onyx in #4976
- Remove hardcoded image extraction flag for PDFs by @emerzon in #4970
- fix: Add type ignore for dynamic sqlalchemy class by @raunakab in #4979
- mypy fix by @Orbital-Web in #4980
- drive service account shared fixes by @evan-onyx in #4977
- Fix migration by @Weves in #4982
- remove check for folder assistant before uploading by @Subash-Mohan in #4975
- fix: Fix bug with incorrect model icon being shown by @raunakab in #4986
- fireflies metadata update by @Orbital-Web in #4993
- Re-adding .epub file support by @emerzon in #4989
- Bugfix/drive doc ids by @evan-onyx in #4990
- Bugfix/drive doc ids3 by @evan-onyx in #4998
New Contributors
- @CaptainJeff made their first contribution in #4806
- @maver1ck made their first contribution in #4699
- @rexjohannes made their first contribution in #4778
- @suvodhoy made their first contribution in #4855
- @ZhipengHe made their first contribution in #4625
- @Django149 made their first contribution in #4869
Full Changelog: v0.29.4...v1.0.0