Natural language to SQL for SQLite

Connect SQLite and ask your first question

Install the package, register the connection once, build the knowledge base, and query. Leave out --password and it is prompted for without echoing, then stored in your OS keychain rather than the connectors file. Only where no keychain is available does it go into the stored connection URL, with a warning.

$ pip install nlqueries-core
$ nlqueries connect sqlite --database /data/app.db --alias app-local
$ nlqueries process-history app-local --annotate
$ nlqueries export-kb app-local
$ nlqueries query app-local "how many orders were placed last month by first-time customers?"

How NL2SQL works on SQLite

Most text-to-SQL tools hand the LLM a schema dump and hope. NLQueries builds a YAML knowledge base from your SQLite schema and from real query history, in this case none; SQLite is file-based and keeps no query history, so it learns which joins your team actually uses and what your columns mean. That knowledge base is what the model sees when it writes SQL.

Every generated statement is parsed with sqlglot and refused unless it is a single read-only query, and its columns are checked against the schema in the knowledge base. It then runs read-only where the engine allows it, with a statement timeout; what each connector enforces has the details per engine. Use nlqueries ask to see the SQL without running it, or nlqueries query to run it. A semantic cache answers repeated questions without another LLM call, and runs the stored SQL again so the rows are current.

SQLite-specific details

SQLite ships with Python, so no extra install is needed. --database is a file path, or :memory: for a transient database.

The knowledge base is built from PRAGMA table_info and PRAGMA foreign_key_list, so declared foreign keys turn directly into join hints.

The file is opened read-only (mode=ro), so writes and DDL are refused, and an authorizer refuses ATTACH and any pragma outside an allow-list. A path that does not exist is refused rather than created. SQLite has no server-side statement timeout, so a runaway query is interrupted best-effort by a watchdog.

Full connector notes are in the connectors guide, and the read-only role to create is in database hardening. This connector covers SQLite files, including app databases and exported datasets.

Use it from Claude, Cursor or any MCP client

Run nlqueries mcp-server start and point Claude Desktop, Cursor, or any Model Context Protocol client at it. Its query tool generates SQL through the same pipeline, runs it read-only and returns the rows. Over stdio, the transport desktop clients use, there is nothing to configure. A network transport will not start without authentication, by OIDC or a pre-shared token, and grants per agent and action decide who may run SQL. See MCP authentication.