kennethreitz/records

Health v2 · maintenance only

Adoption not part of the score

7216 stars · 570 forks observed · 2026-10-19

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Records is a Python library for making raw SQL queries to relational databases with a simple, human-friendly interface for accessing results. Built on SQLAlchemy and Tablib, it supports parameterized queries, transactions, and one-line export to CSV, JSON, YAML, HTML, XLS, or Pandas DataFrames.

Use cases

• run raw sql queries from python without an orm
• export database query results to csv or json
• query postgres with simple parameterized sql
• generate reports from database queries
• access query rows as attributes or dict keys
• run sql queries from the command line

When to choose

• you prefer writing plain SQL over ORM abstractions
• you need quick ad-hoc queries with ergonomic result access
• you want to export query results to multiple formats easily
• you want a lightweight wrapper around SQLAlchemy connections

When to avoid

• you need an ORM with models, relationships, and migrations
• you need async database access
• you need advanced connection pooling or sharding features
• your project requires a very actively maintained library

Facets

library · maturity maintenance

database database-driver orm serialization cli databases data-science developer-tools python cross-platform cli sql sqlalchemy tablib raw-sql query-results data-export postgres mysql sqlite command-line

Member repositories

For agents

markdown · JSON · MCP: product_card(name="kennethreitz/records")