Woody 849beb4d4e feat: add LLMClientDP for Deepseek decompose (Phase 6) ... Uses Deepseek's json_object response_format (not json_schema, which Deepseek does not support). Always disables thinking mode. Includes unit tests (12) and acceptance tests (5). Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent) Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>		2026-05-04 14:58:53 +08:00
..
__init__.py	feat: Phase 1.1 project setup with config, database, and models	2026-04-22 16:13:52 +08:00
chunk_highlight_service.py	feat: track highlight generation prompt, response, and timing in history (Phase 5.5)	2026-04-29 11:18:21 +08:00
embedding_client.py	feat(backend): add embedding client and update LLM client	2026-04-23 13:26:43 +08:00
highlight_cache.py	feat: add SQLite highlight cache service (Phase 5.4.3)	2026-04-29 09:26:20 +08:00
history_service.py	feat: track highlight generation prompt, response, and timing in history (Phase 5.5)	2026-04-29 11:18:21 +08:00
llm_client.py	fix: vLLM structured output missing thinking-control extra_body	2026-04-29 21:01:10 +08:00
llm_client_dp.py	feat: add LLMClientDP for Deepseek decompose (Phase 6)	2026-05-04 14:58:53 +08:00
prompt_service.py	feat(prompts): integrate filter_per_subq with PromptService, fix seed bugs, restructure UI	2026-04-27 11:14:27 +08:00
query_decomposer.py	fix: use vLLM-native guided_json for structured output	2026-04-29 16:49:14 +08:00
rag.py	feat: structured LLM output for decompose + citation fuzzy matching (Phase 5)	2026-04-28 15:39:17 +08:00
relevance_filter.py	fix: wrap filter chunks in XML tags for clearer LLM input	2026-04-30 13:59:03 +08:00