Commit graph

340 commits

Author SHA1 Message Date
Timothy Jaeryang Baek
91b6483aa9 refac 2025-09-28 14:46:01 -05:00
Timothy Jaeryang Baek
4f06f29348 refac 2025-09-28 13:22:39 -05:00
Timothy Jaeryang Baek
742e2ff193 refac 2025-09-28 12:42:02 -05:00
Timothy Jaeryang Baek
97dae432ab refac 2025-09-28 12:23:20 -05:00
Timothy Jaeryang Baek
db9d3d386d refac: mcp spec/response handling 2025-09-28 12:22:11 -05:00
Timothy Jaeryang Baek
4997ef2662 refac 2025-09-26 15:57:03 -05:00
Timothy Jaeryang Baek
cda4c95c86 fix: default tool calling 2025-09-26 15:50:16 -05:00
Timothy Jaeryang Baek
680783266c refac/fix: tool response header type check 2025-09-26 15:42:17 -05:00
Timothy Jaeryang Baek
2fa222d00a refac 2025-09-25 15:20:13 -05:00
Tim Jaeryang Baek
cd417ca0ba
Merge pull request #17744 from Classic298/fix-rag-full-context
Fix: Prevent RAG queries when all files are in full context
2025-09-25 11:55:41 -05:00
google-labs-jules[bot]
4a7e1b93e5 Fix: Prevent RAG queries when all files are in full context
This commit fixes an issue where Retrieval-Augmented Generation (RAG)
queries were still being generated even when all attached files were set
to 'full context' mode. This was inefficient as the full content of the
files was already available to the model.

The `chat_completion_files_handler` in `backend/open_webui/utils/middleware.py`
has been updated to:
- Check if all attached files have the `context: 'full'` property.
- Skip the `generate_queries` step if all files are in full context mode.
- Pass a `full_context=True` flag to the `get_sources_from_items`
  function to ensure it fetches the entire document content instead of
  performing a vector search.

This change ensures that RAG queries are only generated when necessary,
improving the efficiency of the system.
2025-09-25 15:54:58 +00:00
Timothy Jaeryang Baek
77e971dd9f feat: oauth2.1 mcp integration 2025-09-25 01:49:16 -05:00
Timothy Jaeryang Baek
f096e99059 refac: rag context handling 2025-09-24 09:38:14 -05:00
Timothy Jaeryang Baek
1c64269387 refac 2025-09-23 16:51:31 -05:00
Timothy Jaeryang Baek
c55afc4255 refac 2025-09-23 03:19:36 -04:00
Timothy Jaeryang Baek
f1bbf3a91e refac 2025-09-23 03:05:38 -04:00
Timothy Jaeryang Baek
de7f7b3d85 refac 2025-09-23 02:40:59 -04:00
Timothy Jaeryang Baek
777e81f7a8 feat: experimental mcp support 2025-09-23 02:03:26 -04:00
Timothy Jaeryang Baek
60db9ec8ef refac: built-in tools ui component support 2025-09-19 01:38:44 -05:00
Timothy Jaeryang Baek
e4c4ba0979 fix: oauth token 2025-09-19 00:10:48 -05:00
Timothy Jaeryang Baek
7be5b7f50f refac 2025-09-18 22:03:14 -05:00
Timothy Jaeryang Baek
07c5b25bc8 feat: tool ui element support 2025-09-18 20:55:23 -05:00
Timothy Jaeryang Baek
0cfc8f5256 refac 2025-09-17 09:19:56 -05:00
Timothy Jaeryang Baek
c2f98a4cd2 refac: usage info 2025-09-16 10:28:25 -05:00
Timothy Jaeryang Baek
aa8ab349ed feat: ref chat 2025-09-14 10:26:46 +02:00
Timothy Jaeryang Baek
6c0a5fa91c refac 2025-09-11 21:56:59 +04:00
Timothy Jaeryang Baek
cf72f5503f refac 2025-09-11 12:26:25 +04:00
Timothy Jaeryang Baek
b5bb6ae177 refac 2025-09-08 18:50:23 +04:00
Timothy Jaeryang Baek
f71834720e refac 2025-09-08 18:35:09 +04:00
Timothy Jaeryang Baek
217f4daef0 feat: server-side OAuth token management system
Co-Authored-By: Classic298 <27028174+Classic298@users.noreply.github.com>
2025-09-08 18:05:43 +04:00
Timothy Jaeryang Baek
6dc0df2473 refac 2025-09-07 05:17:38 +04:00
Timothy Jaeryang Baek
3d37e4a42d refac 2025-09-07 05:06:03 +04:00
Timothy Jaeryang Baek
1cdb7aed1e refac/enh: status history 2025-09-07 04:21:46 +04:00
Timothy Jaeryang Baek
0a85dd4bca refac: web search link display 2025-09-07 03:00:28 +04:00
Timothy Jaeryang Baek
1bc9711afd refac 2025-09-05 17:28:05 +04:00
Timothy Jaeryang Baek
37bf0087e5 refac: tool message format 2025-09-03 13:57:14 +04:00
Timothy Jaeryang Baek
e830b4959e enh: llama cpp timing stats 2025-09-01 22:49:49 +04:00
Timothy Jaeryang Baek
f56889c5c7 fix: fillter exception handling 2025-09-01 14:14:20 +04:00
Timothy Jaeryang Baek
e0ab5adb97 refac 2025-08-31 23:52:50 +04:00
Timothy Jaeryang Baek
e7c7c65227 refac/fix: error message 2025-08-28 02:35:29 +04:00
Timothy Jaeryang Baek
e39ce16a86 enh: custom reasoning tags 2025-08-27 17:24:16 +04:00
Timothy Jaeryang Baek
31485835a7 enh: query caching
Co-Authored-By: Jacob Leksan <63938553+jmleksan@users.noreply.github.com>
2025-08-27 03:07:21 +04:00
Timothy Jaeryang Baek
e4b6855984 enh: CHAT_RESPONSE_MAX_TOOL_CALL_RETRIES 2025-08-27 02:58:25 +04:00
Timothy Jaeryang Baek
43b2eca418 refac 2025-08-22 14:01:57 +04:00
Timothy Jaeryang Baek
094a16ab49 refac 2025-08-21 03:38:26 +04:00
Timothy Jaeryang Baek
d6f709574e refac/enh: async process chat handling 2025-08-19 01:24:53 +04:00
Timothy Jaeryang Baek
f592748011 refac: tool server redis cache 2025-08-18 20:53:46 +04:00
Timothy Jaeryang Baek
d7363fd65f fix: arena model selected model id 2025-08-17 04:55:45 +04:00
Timothy Jaeryang Baek
ccd2a0be5b refac 2025-08-17 04:15:13 +04:00
Tim Jaeryang Baek
47560d4d72
Merge pull request #14703 from rragundez/code-interpreter-blacklist
feat: Blacklist modules from arbitrary code execution in code interpreter
2025-08-17 04:12:12 +04:00