Lexery - Branch Divergence

Why This Matters

Одна з найважливіших truths про сучасний Lexery: проект існує одночасно в двох активних гілках розвитку.

Exact Observed Snapshot

Local branch vs remote brain branch

  • Local branch: legal-agent-brain-dev
  • Relative to origin/legal-agent-brain-dev: 0 remote-only commits, 10 local-only commits

Local branch vs remote default branch

  • Relative to origin/dev: 21 remote-only commits, 26 local-only commits

What Lives On origin/dev

  • shared contracts package
  • auth refactor
  • auth screens
  • richer profile/user metadata shape
  • subscription plan UI
  • workspace/sidebar/product-shell refinements
  • open auth registration infra PR
  • bounded legal orchestrator
  • clarification pause/resume hardening
  • deterministic ORCH cost cuts
  • DocList recovery
  • brain-admin proposal queue wiring
  • retry reuse and tail-latency tuning

Best Interpretation

  • Observed: origin/dev is the product integration branch.
  • Observed: legal-agent-brain-dev is the runtime/brain innovation branch.
  • Inferred: these are not trivial parallel features; they are partially different product realities.

Consequences

Positive

  • The team can move fast on product shell and brain quality independently.

Negative

  • Root README can drift badly.
  • Contracts can diverge.
  • Plan codes and feature assumptions can split.
  • Integration debt compounds silently.

Documentary Drift Caused By This Split

  • Root README.md still says apps/portal is a placeholder.
  • Current code proves apps/portal is already real.
  • Some portal docs lag behind actual auth/chat/product progress.
  • Azure docs describe a deployment target that current live code only partially proves.

Strategic Reading

Lexery’s current challenge is not lack of architecture. It is reconciling:

  • a fast-moving legal brain branch
  • a fast-moving product shell branch
  • and the contracts that must sit between them

See Also