Template Hub — source × sink matrix

2 source templates × 4 sink templates. ✓ = every stream has a write mode the sink supports; ◐ = some streams do; — = none. Each source section below carries the copy-paste command for every compatible sink.

source \ sink	faucet-hq/bigquery	faucet-hq/jsonl	faucet-hq/postgres	faucet-hq/sqlite
faucet-hq/example-csv	✓	✓	✓	✓
faucet-hq/example-rest-api	✓	✓	✓	✓

Sinks

sink: faucet-hq/bigquery

Google BigQuery — one table per stream, atomic overwrite or keyed MERGE upsert

• connector: bigquery · write modes: append, upsert, delete, overwrite
• params:
  • bq_dataset (default "raw") — Dataset every stream’s table lands in
  • bq_project (required) — GCP project id that owns the dataset
  • bq_sa_key (required, secret) — Service-account key JSON, inline. Pass it with —param-env from a secret store; copy this template and switch auth to application_default to use ADC instead.

sink: faucet-hq/jsonl

Local JSON Lines files, one per stream — the local validation destination

• connector: file · write modes: append, overwrite
• params:
  • out_dir (default "./out") — Directory to write /.jsonl under

sink: faucet-hq/postgres

PostgreSQL — one auto-mapped table per stream, transactional overwrite or ON CONFLICT upsert

• connector: postgres · write modes: append, upsert, delete, overwrite
• params:
  • pg_schema (default "public") — Schema every stream’s table is created in
  • pg_url (required, secret) — postgres://user:pass@host:5432/db connection URL

sink: faucet-hq/sqlite

Local SQLite database — one auto-mapped table per stream, real overwrite/upsert semantics without infrastructure

• connector: sqlite · write modes: append, upsert, delete, overwrite
• params:
  • sqlite_path (default "./out/faucet.db") — Database file (created if missing)

Sources

faucet-hq/example-csv

Example — two CSV exports as two streams (runs offline, no credentials)

• tags: example, file
• connector: file · 2 stream(s)
• params:
  • data_dir (default "./hub/examples/data") — Directory holding orders.csv and customers.csv

stream	write (preference)	primary keys
orders	overwrite → upsert	id
customers	overwrite → upsert	id

→ faucet-hq/bigquery

faucet run --source faucet-hq/example-csv --sink faucet-hq/bigquery \
  --param bq_project=<bq_project> \
  --param bq_sa_key

→ faucet-hq/jsonl

faucet run --source faucet-hq/example-csv --sink faucet-hq/jsonl

→ faucet-hq/postgres

faucet run --source faucet-hq/example-csv --sink faucet-hq/postgres \
  --param pg_url

→ faucet-hq/sqlite

faucet run --source faucet-hq/example-csv --sink faucet-hq/sqlite

faucet-hq/example-rest-api

Example — a bearer-authenticated REST API with cursor pagination and one stream per endpoint

• tags: example, rest
• connector: rest · 2 stream(s)
• params:
  • api_token (required, secret) — Bearer token
  • base_url (required) — API root, e.g. https://api.example.com/v1

stream	write (preference)	primary keys
accounts	overwrite → upsert	id
events	upsert → append	id

→ faucet-hq/bigquery

faucet run --source faucet-hq/example-rest-api --sink faucet-hq/bigquery \
  --param api_token \
  --param base_url=<base_url> \
  --param bq_project=<bq_project> \
  --param bq_sa_key

→ faucet-hq/jsonl

faucet run --source faucet-hq/example-rest-api --sink faucet-hq/jsonl \
  --param api_token \
  --param base_url=<base_url>

→ faucet-hq/postgres

faucet run --source faucet-hq/example-rest-api --sink faucet-hq/postgres \
  --param api_token \
  --param base_url=<base_url> \
  --param pg_url

→ faucet-hq/sqlite

faucet run --source faucet-hq/example-rest-api --sink faucet-hq/sqlite \
  --param api_token \
  --param base_url=<base_url>