For AI agents: the complete documentation index is available at /llms.txt, the full documentation bundle is available at /llms-full.txt, and this page is available as Markdown at /packs/recommended/python.md.
Quick start

Recommended Python

Use this set for Python projects built around pure transformations, immutable domain values, explicit capabilities, and typed Results. It uses returns for failure and composition, with shared rules for engineering decisions and collaboration.

Pack: bytefarm/recommended-python · Version: 0.1.0

Install

With the npm CLI installed, run from a project beside a checkout of thebytefarm/granary named granary-packs. The checkout supplies the pack sources; no build is needed:

granary add file:../granary-packs/packs#recommended-python --output .granary/context/docs/bytefarm

Included packs

This composite installs these five packs in reading order. Open a card to browse that pack's concepts and rules.

Use the installed context

Follow the reference links in root AGENTS.md. Read the shared engineering rules first, then the topics relevant to your change. Python rules apply to Python source and examples. The documented returns decorator and composition typing uses its compatible mypy plugin; adding dependencies still follows your project's approval rules.

Adding another recommended language set reuses the shared packs. The lockfile pins the source revision and file integrity; granary install restores the same context.