v0.4.4 — Citation Parsing & Mind Map Fix

What's Changed

🐛 Fixed

• Complex citation parsing (PR #84) — Fixed notebook_query dropping cited text when Google returns "direct" citation segments (integer-first elements) alongside the standard "wrapped" format. Both segment variants are now correctly handled. Thanks to @meirtsvi for this contribution!

• Table citation extraction (PR #84) — When a citation references a data table, the response now includes a structured cited_table field with num_columns and rows data. Table segments are indicated by a <cited_table> placeholder in the cited_text field.

• Mind map JSON missing from MCP response (Issue #83) — studio_create for mind maps was returning metadata (root_name, children_count) but dropping the actual mind_map_json. The full JSON now flows through to MCP clients. Thanks to @cowhi for reporting!

✅ Added

• 17 new unit tests for citation parsing — Comprehensive test coverage for direct/wrapped segment detection, table placeholder insertion, table row parsing, and table extraction from detail structures. Total tests: 503.

Install / Upgrade

# PyPI
pip install --upgrade notebooklm-mcp-cli

# or with uv
uv tool install --force notebooklm-mcp-cli

Full Changelog: v0.4.3...v0.4.4