v0.6.0

New Features:

1. Context-Aware Chunk Enrichment
   • Chunk enrichment now uses cached document summaries when available.
   • Added surrounding Markdown context before and after each chunk to improve inferred titles, summaries, keywords, questions, and context fields.

2. Bulk Chunk Enrichment
   • Added sidebar bulk chunk enrichment for multiple documents.
   • Reuses matching saved chunks when available; otherwise chunks first, enriches, and saves the enriched chunk set.
   • Added a setting to choose whether bulk chunking/enrichment uses the first available Markdown or the converter selected in settings.

Improvements:

1. Bug Fixes & Refactoring
   • Tightened backend Markdown/document matching and filename validation.
   • Improved chunk-context safety to avoid using stale or incorrect offsets.
   • Hardened chunk normalization on both backend and frontend.
   • Fixed frontend stream handling, stale document response guards, and MD-to-PDF filename selection.
   • Removed dead code and unused imports.

2. Documentation & UX Polish
   • Updated README to reflect the stable chunk enrichment flow.
   • Removed beta wording from chunk enrichment.
   • Refined the feature list and enrichment description for the 0.6.0 release.
   • Fixed a frontend CSS syntax issue during build.