TypeScript + Go + Rust 3,100 FILES 5 FINDINGS

HOW BAML
ACTUALLY WORKS

A programming language to get structured data from LLMs — build reliable AI applications. Conventions, patterns, and architecture extracted from the boundaryml/baml repository by sourcebook.

VIEW_REPO north_east
QUICK_REF:
Testing: Vitest + co-located State: Jotai atoms

WHAT_MATTERS

  • →

    This is a monorepo. Changes may affect multiple packages. Check workspace dependencies before modifying.

KEY_FINDINGS

This is a monorepo. Changes may affect multiple packages. Check workspace dependencies before modifying.

HIGH

Generated files detected — do NOT edit these directly. Modify the source/schema they are generated from.

HIGH

State management uses Jotai atoms. Follow this pattern for new data operations.

HIGH

Uses barrel exports (index.ts files that re-export). Import from the directory, not from deep paths.

HIGH
+ 1 MORE FINDINGS (MEDIUM CONFIDENCE)

5 reverted commits including "implement never type". Check anti-patterns.

MED
GENERATED IN ~3 SECONDS WITH

npx sourcebook init

VIEW ON GITHUB star

FROM_THE_BLOG

Why AI agents fail on real codebases arrow_forward Why auto-generated context makes agents worse arrow_forward

RELATED_REPOS

FERN