CSV connector

Runs SQL over CSV files on an embedded DuckDB — no database to stand up. Each file in the directory becomes a queryable view named after the file (sales.csv → sales). This is what these docs use.

# sources.yaml
main:
  type: csv
  directory: data        # folder of .csv files, relative to the project

Then query the view by file name:

SELECT month, SUM(downloads) AS downloads
FROM downloads            -- data/downloads.csv
GROUP BY month

Key	Purpose
`directory`	Folder of CSV files (each → a view).
`files`	Or an explicit `{view_name: path}` map.
`path`	A single CSV file.

Extra: none — it's in the core install. Inherits the DuckDB connector's reconnect-on-fatal resilience.