This release includes 1 breaking change for platform teams planning a safe upgrade.
✓ No known CVEs patched in this version
Topics
+1 more
ReleasePort's take
Moderate signalThis release fixes several bugs affecting label search, XML handling, LaTeX escaping, and error mapping in the pipeline. It also deprecates the docs_format="lightrag" enqueue entrypoint.
Why it matters: Bugfixes improve stability of Neo4j label searches, multimodal pipelines, and LLM JSON processing; the docs_format="lightrag" removal triggers a severity‑70 trigger requiring migration before use.
Summary
AI summaryUpdates What's New, pipeline, and parser across a mixed release.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Low |
Standardize parser engine interface through a registration mechanism supporting third‑party parsers. Standardize parser engine interface through a registration mechanism supporting third‑party parsers. Source: llm_adapter@2026-06-11 Confidence: high |
— |
| Deprecation | High |
Remove the deprecated docs_format="lightrag" enqueue entrypoint from pipeline. Remove the deprecated docs_format="lightrag" enqueue entrypoint from pipeline. Source: llm_adapter@2026-06-11 Confidence: high |
— |
| Bugfix | Medium |
Sanitize Lucene reserved chars in neo4j label search. Sanitize Lucene reserved chars in neo4j label search. Source: llm_adapter@2026-06-11 Confidence: high |
— |
| Bugfix | Medium |
Strip XML-illegal control chars before they reach the graph in multimodal processing. Strip XML-illegal control chars before they reach the graph in multimodal processing. Source: llm_adapter@2026-06-11 Confidence: high |
— |
| Bugfix | Medium |
Repair LaTeX escape damage in LLM JSON responses for multimodal analysis and entity extraction. Repair LaTeX escape damage in LLM JSON responses for multimodal analysis and entity extraction. Source: llm_adapter@2026-06-11 Confidence: high |
— |
| Bugfix | Medium |
Map parse-stage failures onto the legacy error-document UI fields in pipeline processing. Map parse-stage failures onto the legacy error-document UI fields in pipeline processing. Source: llm_adapter@2026-06-11 Confidence: high |
— |
Full changelog
What's New
- feat(parser): standardize the parser engine interface through a registration mechanism with support for third-party parsers by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3235
What's Changed
- 🐛 fix(neo4j): sanitize Lucene reserved chars in label search by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3233
- fix(multimodal): strip XML-illegal control chars before they reach the graph by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3237
- fix: repair LaTeX escape damage in LLM JSON responses (multimodal analysis & entity extraction) by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3238
- fix(pipeline): map parse-stage failures onto the legacy error-document UI fields by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3239
- refactor(pipeline)!: remove the deprecated docs_format="lightrag" enqueue entrypoint by @danielaskdd in https://github.com/HKUDS/LightRAG/pull/3240
Full Changelog: https://github.com/HKUDS/LightRAG/compare/v1.5.1...v1.5.2
Breaking Changes
- Removed deprecated docs_format="lightrag" enqueue entrypoint in pipeline
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
About LightRAG
[EMNLP2025] "LightRAG: Simple and Fast Retrieval-Augmented Generation"
Related context
Earlier breaking changes
- v1.5.3 One-time automatic migration copies legacy collection to model‑suffixed schema on upgrade.
- v1.5.0rc3 Removes deprecated `QueryParam.model_func` field from lightrag/base.py.
- v1.5.0rc3 Deprecates the `ENTITY_TYPES` environment variable; replace with ENTITY_TYPE_PROMPT_FILE.
- v1.5.0rc2 ENTITY_TYPES environment variable deprecated; replace with ENTITY_TYPE_PROMPT_FILE before upgrading.
Beta — feedback welcome: [email protected]