XERJ v1.0.0-rc.95 Adds Legacy Office Support

This is a focused but valuable format-coverage upgrade for agent-oriented search, especially where enterprise archives still contain pre-2007 Office files.

• –Legacy Word and Excel files can now enter the same zero-config indexing path as code, PDFs, logs, and modern documents.
• –Developers avoid maintaining separate conversion or extraction pipelines for older archives.
• –The release strengthens XERJ’s fit for RAG, security audits, migrations, and agent memory across messy real-world datasets.
• –The update is incremental and still marked release candidate, so production users should validate parsing against their own Office corpus.
• –XERJ’s broader pitch remains a local Rust binary with Elasticsearch compatibility and agent-focused retrieval. [Product documentation](https://xerj.org/)