RAG-Anything

v1.3.1 Bugfix

This release fixes issues for SREs watching stability and regressions.

Published 13d RAG & Retrieval

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

multi-modal-rag retrieval-augmented-generation

ReleasePort's take

Moderate signal

editorial:auto 13d

RAG-Anything v1.3.1 fixes duplicate detection for text-bearing `insert_content_list` and parsed document calls by deferring `doc_status` creation until after LightRAG ainsert completes. Multimodal-only content paths remain unchanged.

Why it matters: Text-bearing content insertion was creating duplicates due to premature doc_status initialization. Update to 1.3.1 if using insert_content_list with text; ensures data integrity across document processing flows.

Summary

AI summary

Fixed duplicate detection for text-bearing insert_content_list calls and parsed documents by deferring doc_status creation.

Changes in this release

Type	Severity	Summary	CVE
Feature	Medium	Apply the same deferral to process_document_complete for parsed documents with text. Apply the same deferral to process_document_complete for parsed documents with text. Source: llm_adapter@2026-05-21 Confidence: high	—
Feature	Medium	Add regression coverage for content-list and parsed-document duplicate handling. Add regression coverage for content-list and parsed-document duplicate handling. Source: llm_adapter@2026-05-21 Confidence: low	—
Performance	Medium	Run validation commands for code quality and testing. Run validation commands for code quality and testing. Source: llm_adapter@2026-05-21 Confidence: low	—
Bugfix	Medium	Fix duplicate detection for text-bearing insert_content_list calls by deferring doc_status creation until after LightRAG ainsert runs. Fix duplicate detection for text-bearing insert_content_list calls by deferring doc_status creation until after LightRAG ainsert runs. Source: llm_adapter@2026-05-21 Confidence: high	—
Bugfix	Medium	Keep early doc_status creation for multimodal-only content that does not call ainsert. Keep early doc_status creation for multimodal-only content that does not call ainsert. Source: llm_adapter@2026-05-21 Confidence: high	—

Full changelog

What's Changed

Fix duplicate detection for text-bearing insert_content_list calls by deferring doc_status creation until after LightRAG ainsert runs.
Apply the same deferral to process_document_complete for parsed documents with text.
Keep early doc_status creation for multimodal-only content that does not call ainsert.
Add regression coverage for content-list and parsed-document duplicate handling.

Validation

PYTHONPATH=. .venv/bin/python -m pytest tests/test_insert_content_list.py tests/testparser_wiring.py tests/testparser_kwargs.py -q
uvx ruff check raganything/processor.py tests/test_insert_content_list.py raganything/__init__.py
.venv/bin/python -m compileall raganything/processor.py tests/test_insert_content_list.py raganything/__init__.py

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track RAG-Anything

Get notified when new releases ship.

About RAG-Anything

"RAG-Anything: All-in-One RAG Framework"

All releases →