Commit Graph

34 Commits

Author SHA1 Message Date
f58bced86e chore(prompt): improve layout instructions for Chinese 2025-02-25 15:58:37 +08:00
f077807128 feat(ai): improve error handling 2025-02-24 20:06:04 +08:00
ac716043a2 fix(deep-research): context size in settings are not applied to trimPrompt 2025-02-24 16:58:43 +08:00
66c28978d1 feat(search): retry failed nodes 2025-02-22 15:14:12 +08:00
7b1bcfc408 fix(deep-research): pass search results incrementally when triggering node_complete 2025-02-21 23:31:24 +08:00
5d1901f9a4 fix(deep-research): parent nodes are accidently triggered twice ingenerating_query 2025-02-21 23:30:10 +08:00
a75539d481 fix(deep-research): JSON reserved words should be escaped 2025-02-20 14:54:08 +08:00
f593546864 fix(deep-research): searchLanguage is missing when recursively calling deepResearch 2025-02-18 12:57:56 +08:00
fc83d9387e feat: support setting concurrency limits in web search 2025-02-17 16:20:25 +08:00
c10090d7d9 feat: rewritten search visualization using VueFlow 2025-02-16 21:40:35 +08:00
3830875858 feat(DeepResearch): show URL title in "Visited URLs" if possible 2025-02-15 21:48:35 +08:00
f8af8b4afc feat: add support for Firecrawl 2025-02-15 21:41:15 +08:00
41abca144d fix(feedback.ts): throw error without wrapping 2025-02-15 20:35:37 +08:00
970ca029f4 feat: support DeepSeek, OpenRouter & Ollama providers 2025-02-15 15:31:25 +08:00
16291788e4 fix: GPT-4o returns search query undefined in certain cases 2025-02-14 19:40:42 +08:00
e15e35e5dd fix: remove timeout limits in processSearchResult 2025-02-14 16:13:41 +08:00
e7296df78f feat: support reasoning models like DeepSeek R1 2025-02-14 15:20:02 +08:00
e60890e286 feat: improve error handling in web search 2025-02-14 10:30:02 +08:00
83f64b5948 feat: search in a different language 2025-02-13 22:26:43 +08:00
1ce32fabc2 feat: improve error handling 2025-02-13 16:46:11 +08:00
7ef3fe3492 fix: revert maxResults in deep-research.ts 2025-02-13 00:24:12 +08:00
f1dd76b297 fix: "export as .pdf" broken 2025-02-13 00:08:49 +08:00
b0c86ce2e2 feat: respond in user's language, style fixes and research depth fixes 2025-02-12 19:34:21 +08:00
5a973b5d63 fix(feedback): force the AI model to generate x questions instead of "max x" 2025-02-12 17:01:03 +08:00
9f2760d8b8 chore: remove lodash 2025-02-12 15:46:11 +08:00
6d6124d4ef feat: i18n support 2025-02-12 15:29:26 +08:00
a7e4124620 build: use manual chunks to reduce dist size 2025-02-12 10:27:08 +08:00
4b75c254a5 feat: config manager modal 2025-02-12 00:05:27 +08:00
d027965013 style: prettier format 2025-02-11 22:57:49 +08:00
84f63abb3d refactor: make requests purely client-side 2025-02-11 22:28:58 +08:00
8c81b9a425 feat: generate research report 2025-02-11 21:07:40 +08:00
e971a61bd3 chore: cleanup unused ResearchStep params 2025-02-11 19:26:45 +08:00
2c4eaed03c feat: Deep Research section 2025-02-11 19:16:07 +08:00
2cbd20a1da init 2025-02-11 09:05:47 +08:00