lxfight e2f3b00088 fix(kb): clean up KBMedia records on document delete, fix update_kb_stats (#9120) ...

* fix(kb): clean up KBMedia records on document delete, fix update_kb_stats

- delete_document_by_id now deletes associated KBMedia records before
  removing the KBDocument row, preventing orphaned media data.
- update_kb_stats now passes kb_id filter to vec_db.count_documents,
  fixing chunk count that previously counted across all KBs.
- add tests for document deletion cleanup and update_kb_stats behavior.

* fix(kb): merge media and document deletion into single transaction

2026-07-03 22:13:23 +08:00

..

chunking

feat(kb): add Markdown-aware chunker for structured documents (#8151)

2026-05-30 17:43:13 +08:00

parsers

fix: support rst and adoc knowledge uploads (#8255)

2026-05-20 14:38:16 +08:00

retrieval

fix(kb): improve retrieval resilience on per-KB failure and rerank selection (#9122)

2026-07-03 22:12:24 +08:00

kb_db_sqlite.py

fix(kb): clean up KBMedia records on document delete, fix update_kb_stats (#9120)

2026-07-03 22:13:23 +08:00

kb_helper.py

fix: paginate knowledge base dashboard lists (#9055)

2026-06-28 14:00:45 +08:00

kb_mgr.py

fix(kb): replace fragile error string matching with pre-check for duplicate kb_name (#9121)

2026-07-03 22:12:00 +08:00

models.py

style: add ruff lint module of isort and pyupgrade, and some ruff check fix (#3214)

2025-11-01 13:26:19 +08:00

prompts.py

feat: Add URL document parser for knowledge base (#3622)

2025-11-17 19:05:14 +08:00