JSON connector

Runs SQL over JSON files on an embedded DuckDB — no database to stand up. Each file in the directory becomes a queryable table named after the file (orders.json → orders), via DuckDB's read_json_auto, which auto-detects both a JSON array of objects and newline-delimited JSON. So .json, .ndjson, and .jsonl files are all picked up.

# sources.yaml
main:
  type: json
  directory: data        # folder of .json/.ndjson/.jsonl files, relative to the project

Then query the table by file name:

SELECT region, SUM(amount) AS revenue
FROM orders               -- data/orders.json
GROUP BY region

Key	Purpose
`directory`	Folder of JSON files (each → a table by stem).
`files`	Or an explicit `{table_name: path}` map.

Extra: none — it's in the core install (the json reader ships in core DuckDB). Inherits the DuckDB connector's reconnect-on-fatal resilience, and dashdown query --tables / --schema <table> work out of the box.