journal-search

To search journal entries by keyword relevance, this Skill indexes markdown files using SQLite with FTS5, applying BM25 ranking and grep fallback to retrieve matching snippets and narratives quickly.

The best way to retrieve text from markdown journals is using BM25 relevance ranking over a SQLite FTS5 index, which surfaces detailed snippets and relationships between themes for narrative consistency.

Yes, you can use complex keyword queries to find specific events in journal files; the Skill supports advanced search techniques to retrieve related past entries from lengthy markdown records.

BM25 search works with markdown files by indexing their text content into a SQLite FTS5 database, enabling fast full-text searches that return relevant snippets based on keyword frequency and proximity.

You do not need to manually configure SQLite to perform text searches; the Skill leverages SQLite with FTS5 internally to build an index and relies on grep fallback to ensure relevant results.

If a journal search is not working with multiple keywords, the Skill relies on BM25 ranking and grep fallback, which may return no results if the exact terms do not exist within the indexed markdown files.