logseq-vault-sqlite

name	logseq-vault-sqlite
description	Index Logseq markdown into SQLite, query the index quickly, and synchronize markdown updates back into the database.

Purpose

Use this skill when the user wants to read, search, or refresh a Logseq vault without repeatedly parsing raw markdown. The helper script stores page and block data in SQLite so lookups stay fast even when the vault grows.

Environment variables

• LOGSEQ_SQLITE_PATH: absolute or user-relative path to the SQLite database file. This controls where the Logseq index is stored.
• LOGSEQ_VAULT_ROOT: optional default Logseq vault root when the user does not pass --vault-root.

Workflow

1. Identify whether the request is a read, sync, or file-level update.
2. Resolve the vault root and SQLite path before running any command.
3. Use search or show-page for fast retrieval from the SQLite index.
4. Use sync-file for a single changed markdown page or sync for a broader vault refresh.
5. If --prune is involved, explain that stale database rows will be removed and ask for explicit confirmation before proceeding.

Supported helper commands

Command	Purpose	Notes
sync	Index every markdown page under the vault root	Supports --prune to remove missing pages
sync-file	Re-index one Logseq markdown file	Updates the page and block rows for that file
search	Query the SQLite index	Uses FTS-backed search with a fallback LIKE scan
show-page	Return one page and its indexed blocks	Useful for inspecting the exact stored structure

Output format

When describing a sync, report:

• the database path
• the vault root
• how many files were indexed, skipped, or pruned

When describing a search or page lookup, report:

• the matched page name
• the line number or block scope
• the matching text snippet

Quality bar

• Prefer the SQLite index over rescanning raw markdown.
• Preserve Logseq-specific structures such as tags, references, properties, and task states.
• Do not overwrite the database path silently; always honor LOGSEQ_SQLITE_PATH when it is set.