gaya 0.1.0

Simple data quality checks that just work.

Gaya

Simple data quality checks that just work.

pip install gaya
gaya init
gaya run

---

What it checks

Check	Default
Null rate per column	warn >10%, fail >25%
Required columns	zero nulls
Primary key uniqueness	zero duplicates
Row count change	warn >20%, fail >40%
Schema drift	warn on add, fail on remove

Quick config

datasources:
  main_db:
    type: postgres
    host: localhost
    database: app_db
    user: app_user
    password: env:DB_PASSWORD

tables:
  orders:
    source: main_db
    layer: staging
    primary_key: order_id
    not_null:
      - order_id
      - customer_id

Output

  ──────────────────────────────────────────────────────
  ✖  staging.orders  FAILED
     ✖  row count dropped 38% (1.2M → 740K)
          → A drop this large usually means a failed upstream load.

  ──────────────────────────────────────────────────────
  1 table(s) · 1 failed · 7 passed
  Finished in 2.3s
  ──────────────────────────────────────────────────────

Exit codes (CI-friendly)

Code	Meaning
0	All passed
1	Warnings only
2	One or more failures
3	Gaya error (config / connection)

CI integration

# GitHub Actions
- name: Run data quality checks
  run: gaya run --quiet

---

Free for local databases (CSV, Postgres, MySQL).
Snowflake connector → #\x00 \x00g\x00a\x00y\x00a\x00_\x00d\x00q\x00 \x00 \x00