Skip to content

decisions.md

Latest commit

 

History

History
26 lines (21 loc) · 1.44 KB

decisions.md

File metadata and controls

26 lines (21 loc) · 1.44 KB

Design Decisions

Decision 1: Keep FastAPI app structure stable

  • Choice: preserve existing app/ runtime layout to avoid regressions.
  • Why: project already has end-to-end behavior and domain logic coupled to templates/connectors.
  • Tradeoff: full package migration is incremental rather than immediate.

Decision 2: Add typed public package for reusable scoring

  • Choice: introduce src/operational_leverage_framework (internal package name) with explicit core/io/models/cli.
  • Why: provide a clean public API boundary for portfolio usage and offline demos.
  • Tradeoff: temporary dual-path imports (app/services and package wrappers).

Decision 3: Deterministic offline examples

  • Choice: examples use local JSON evidence inputs and deterministic scoring.
  • Why: reproducible outputs for reviewers without connector/API dependencies.
  • Tradeoff: examples represent a narrow subset of full workflow complexity.

Decision 4: Quality gates with scoped mypy

  • Choice: enforce Ruff/pytest globally and mypy on typed boundary modules.
  • Why: practical maintainability without blocking on complete typing of all legacy paths.
  • Tradeoff: some deeper modules remain outside strict type coverage.

Decision 5: Standardize local launcher

  • Choice: use python scripts/run.py web as the primary local entrypoint.
  • Why: one predictable cross-platform startup command for GitHub users.
  • Tradeoff: launcher logic is centralized in a Python wrapper script.