sk/Perplexica
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
562 commits 17 branches 33 tags 47 MiB
Commit graph

202 commits

Author SHA1 Message Date
Willie Zutz
e707bc630f feat(agent): Minor tweaks 2025-06-10 00:12:45 -06:00
Willie Zutz
4d85210d50 fix(package): New gemini package signature changed. 2025-06-09 23:08:03 -06:00
Willie Zutz
b4e2585856 feat(agent): Pretty decent first pass at agent mode 2025-06-09 23:00:25 -06:00
Willie Zutz
3f1f437d4f Merge branch 'working' of github.com:boarder2/Perplexica into working 2025-05-31 11:56:59 -06:00
Willie Zutz
6422c5b7e0 fix(api): Allow persona instructions to be passed in the search API 2025-05-31 11:54:52 -06:00
Willie Zutz
48b36d39cc chore: get rid of unused prompt files and make sure persona instructions are used everywhere 2025-05-27 23:51:09 -06:00
Willie Zutz
011d10df29 feat(UI): allow system prompts and persona prompts to be saved server side and individually included or excluded from messages 2025-05-27 12:53:30 -06:00
Willie Zutz
8e6934bb64 feat(prompt): Use date, not time in prompts 2025-05-26 11:05:11 -06:00
Willie Zutz
4edd173207 feat(optimization): Allow retrieving more sources in quality mode if we can get through them quickly 2025-05-25 15:32:47 -06:00
Willie Zutz
a5cd2fa089 feat(UI): More progress detail 2025-05-25 14:51:21 -06:00
Willie Zutz
8ce50b48f0 feat(optimization): improve quality mode performance by limiting source gathering 2025-05-25 02:14:29 -06:00
Willie Zutz
6343dd5303 fix(logging): remove excessive logging 2025-05-24 14:41:33 -06:00
Willie Zutz
87a7ffb445 feat(web): Use crawlee/playwright to retrieve web content in quality mode. It retrieves content more reliably than fetch + JSDoc, at the expense of speed. 2025-05-24 14:37:19 -06:00
Willie Zutz
044f30a547 feat(AI): don't rerank docs based on preview. just rely on search results for relevance 2025-05-23 23:11:28 -06:00
Willie Zutz
c47a630372 feat(app): Introduce quality mode. Improve functionality of balanced mode using readability to get page content and pull relevant excerpts 
feat(UI): Show progress during inferrence
feat(security): Don't show API keys in the UI any more
feat(models): Support Claude 4 Anthropic models
 2025-05-23 18:03:35 -06:00
Willie Zutz
288120dc1d feat(UI): protect API keys by not displaying them in the UI. 2025-05-20 00:44:52 -06:00
Willie Zutz
67287b0604 feat(UI): Always show the selected focus mode 2025-05-19 23:31:29 -06:00
Willie Zutz
db3a97b042 feat(prompt): Make the llm handle site limitations and use site:example.com syntax 2025-05-15 00:33:26 -06:00
Willie Zutz
e4f695ec9b update(README): Updated to reflect features in this fork 
fix(cancel): moved to a lib component
 2025-05-14 13:25:06 -06:00
Willie Zutz
9c7ccf42fc feat(app): allow stopping requests 2025-05-14 12:19:22 -06:00
Willie Zutz
380216e062 feat(editing): Adds the ability to edit messages 2025-05-13 01:41:41 -06:00
Willie Zutz
b96c4234f4 update(autocomplete): removed unecessary API logging 2025-05-11 14:14:35 -06:00
Willie Zutz
c3a1e434e5 feat(opensearch): Add autocomplete functionality to opensearch. This gets proxied to the searxng instance. 2025-05-11 13:44:47 -06:00
Willie Zutz
03b27d9cbb fix(UI): Fix results showing in light mode 
feat(AI): Enhance system prompt for more reliable and relevant results
fix(Reddit): Reddit focus should work again. Works around SearXNG limitations of broken reddit search by using `site:reddit.com`
 2025-05-10 15:09:00 -06:00
Willie Zutz
2a37f672ab feat(UI): Tabbed interface for messages. 2025-05-09 01:13:18 -06:00
Willie Zutz
85605fe166 feat(opensearch): Add BASE_URL config to support reverse proxy deployments 2025-05-08 20:58:04 -06:00
Willie Zutz
d839769d7e fix(opensearch): Implement dynamic OpenSearch XML generation and update layout reference 2025-05-08 00:39:17 -06:00
Willie Zutz
ddfe8c607d feat(search): Implement OpenSearch support 
feat(search): Add searchUrl to message
feat(parsers): Enhance parsers to deal with some thinking models better.
 2025-05-08 00:21:31 -06:00
Willie Zutz
f65b168388 feat(UI): Add the search query to the response. 
Also, tweaked the search retriever prompt so it gives better search queries.
 2025-05-07 01:16:51 -06:00
Willie Zutz
8796009141 fix(api): History rewriting should delete the current message. 
fix(UI): Model changes shouldn't submit the form.
 2025-05-06 23:45:46 -06:00
Willie Zutz
6220822c7c feat(app): Allow selecting the AI model at any time without opening the settings page. 
Allow changing focus mode at any time while chatting.
Styling tweaks.
 2025-05-05 00:05:19 -06:00
Willie Zutz
8241c87784 feat(app): 
Adds auto scrolling.
Adds syntax highlighting to code blocks.
 2025-05-03 16:03:04 -06:00
Willie Zutz
c8def1989a fix(Library): Returns metadata back to original format so sources continue to work. 2025-05-01 12:17:08 -06:00
Willie Zutz
a71e4ae10d feat(app): 
- Adds true chat mode. Moves writing mode to local research mode.
- Adds model stats that shows model name and response time for messages.
- Adds settings toggle to allow turning off automatic suggestions
 2025-05-01 11:32:13 -06:00
Willie Zutz
abf9dbb8ba Merge remote-tracking branch 'upstream/master' 2025-04-29 10:23:52 -06:00
ItzCrazyKns
06ff272541 feat(openai): add GPT 4.1 models 2025-04-29 13:10:14 +05:30
ItzCrazyKns
4154d5e4b1 Merge branch 'pr/629' 2025-04-23 20:35:52 +05:30
Willie Zutz
59a10d7d00 Ran prettier formatting 2025-04-20 13:12:23 -06:00
Willie Zutz
67ee9eff53 Apply context window everywhere. Ensure styling is good on all screen sizes. Cleanup inconsistencies with upstream branch. 2025-04-20 13:10:59 -06:00
Willie Zutz
0bb860b154 Fixes history rewrite bug 2025-04-20 11:57:48 -06:00
Willie Zutz
c0705d1d9e Support for Ollama context window configuration 2025-04-20 01:37:10 -06:00
Willie Zutz
73b5e8832e Removed compact mode 2025-04-19 13:36:50 -06:00
Willie Zutz
b2da9faeed More merge 2025-04-19 12:52:15 -06:00
Willie Zutz
1a2ad8a59d Merge remote-tracking branch 'upstream/master' 2025-04-19 12:51:57 -06:00
ItzCrazyKns
1862491496 feat(settings): add LM Studio API URL 2025-04-12 11:59:05 +05:30
ItzCrazyKns
073b5e897c feat(app): lint & beautify 2025-04-12 11:58:52 +05:30
haddadrm
7e1dc33a08 Implement provider formatting improvements and fix client-side compatibility 
- Add PROVIDER_INFO metadata to each provider file with proper display names
- Create centralized PROVIDER_METADATA in index.ts for consistent reference
- Update settings UI to use provider metadata for display names
- Fix client/server compatibility for Node.js modules in config.ts
 2025-04-11 19:18:19 +04:00
haddadrm
aa240009ab Feature: Add LM Studio provider integration - Added LM Studio provider to support OpenAI compatible API - Implemented chat and embeddings model loading - Updated config to include LM Studio API endpoint 2025-04-11 19:18:19 +04:00
sjiampojamarn
41b258e4d8 Set speech message before return 2025-04-08 23:17:52 -07:00
ItzCrazyKns
da1123d84b feat(groq): update model name 2025-04-07 23:30:51 +05:30