We're pleased to announce Apache Answer 2.0.1, a focused patch release that builds on 2.0.0 with better AI retrieval, smoother plugin development, and a solid round of fixes across uploads, moderation, security, and UI details.
This is a small release, but it improves both day-to-day usage and the reliability of newer 2.0 workflows.
✨ Highlights
Smarter AI Retrieval
-
Semantic search in AI chat (#1510) by @hgaol
AI chat can now use semantic search and embedding capabilities to retrieve more relevant community content. -
Vector search plugin and sync service by @hgaol
This release adds the foundation for vector-based retrieval, including plugin support and sync services for question and answer embeddings.
Better Plugin Development
- Improved local plugin path resolution (#1520) by @hgaol
Local plugin development is now more predictable, especially when resolving local paths and handling module replacement.
Fixes That Matter in Production
- Attachment uploads fixed after upgrading to 2.0.0 (#1527) by @robinv8
- Helm install port aligned with service port (#1522) by @Herrtian
- Avatar field updated to support long URLs (#1499) by @maishivamhoo123
- Better dark mode styling for user bubbles (#1505) by @MakiWinster72
🔒 Reliability and Security Improvements
Alongside the release note items, the code changes between v2.0.0 and v2.0.1 also tighten several important areas:
- AI chat content rendering is handled more safely and correctly.
- Admin moderator visibility checks were improved for timeline objects, answers, and comments.
- API key scope validation was strengthened to improve authorization safety.
- Comment and object ID validation was normalized.
- Image decoding checks were improved for uploaded JPEG, PNG, and GIF files.
- Dynamic email template content now escapes HTML characters correctly.
These changes make 2.0.1 a recommended upgrade for teams already running 2.0.0.
🙌 Thanks to Our Contributors
Thank you to everyone who contributed code, fixes, and review to this release:
@Herrtian, @hgaol, @kumfo, @LinkinStars, @maishivamhoo123, @MakiWinster72, and @robinv8.
We appreciate your help in making Apache Answer more capable and more reliable with every release.