Overview
Consolidate two independent streams of component work into a single coherent component library in ux-prototype:
-
jmpicnic branch (
jmpicnic/list-view-management-620) — entity-data-grid factory, cell atoms (8 types), data-grid molecule, dirty tracking, column persistence. Focus: general-purpose grid infrastructure. -
callil branch (
jmpicnic/callil-consolidation, via worktree) — sidebar, app-header, item-details, item-grid organisms, shadcn/ui primitives layer, design token system. Focus: page-level composition and visual design.
The consolidation is not a mechanical merge. It requires semantic analysis of overlapping code, architectural decisions about component organization, and evolutionary changes to the grid infrastructure to support the capabilities both architects implemented independently.
Key Outcomes
Section titled “Key Outcomes”- Unified component hierarchy — all components under
src/components/canary/with consistent naming, imports, and patterns (primitives → atoms → molecules → organisms). - Entity-data-grid as the grid foundation — item-grid (and future entity grids) built on entity-data-grid, eliminating duplicated framework code.
- Arda design tokens — orange-branded palette, responsive control heights, and themeQuartz AG Grid styling as the visual foundation.
- Published package —
@arda-cards/design-systembarrel exports reflect the consolidated library.
Reader’s Guide
Section titled “Reader’s Guide”Analysis Documents
Section titled “Analysis Documents”Start here to understand the decisions. Each document covers a specific topic with options evaluated, decisions recorded (with dates), and rationale.
| Document | What it covers | Start here if you want to understand… |
|---|---|---|
| analysis/index.md | Master index of all decisions | All cross-cutting decisions in one place — branch strategy, canonical paths, component organization, barrel exports, styles, dependencies |
| analysis/grid-integration.md | Entity-data-grid evolution | How item-grid’s capabilities (auto-publish, search, actions, pagination, theme) are promoted into entity-data-grid |
| analysis/primitives.md | Component organization | Why stock shadcn goes in canary/primitives/ and custom components go in canary/atoms/; includes organism import mapping |
| analysis/styles.md | CSS token system | Why the worktree’s Arda orange palette replaces the main clone’s blue-slate; how tokens.css and globals.css are restructured |
| analysis/dependencies.md | npm package changes | Which worktree dependencies to keep/drop and why; library build analysis for runtime vs dev classification |
| analysis/test-coverage.md | Stories, tests, VRT | What test/story gaps the consolidation creates and how to fill them; play function step DSL convention; VRT expected variances |
| analysis/canary-refactor.md | Validation harness | Why canary-refactor/ is adapted after (not during) consolidation and what deferred sub-tasks exist |
Decision Log
Section titled “Decision Log”| Document | What it covers |
|---|---|
| decision-log.md | Chronological record of all 45 decisions + external tickets, with rationale and source document links |
Project Blueprint and Plans
Section titled “Project Blueprint and Plans”The execution plan. Read this after the analysis to understand the sequencing.
| Document | What it covers |
|---|---|
| project-blueprint.md | 10 runs (0-9) sequenced risk-first with verification gates, intermediate checkpoints, Playwright MCP visual checks, and entry/exit criteria per run |
| plan/team-split-assessment.md | Complexity analysis justifying the 10-run decomposition |
| plan/choreography.md | Execution sequence, artifact dependencies, hand-off protocol, recovery procedures |
| plan/run-N-*/project-plan.md | Detailed task lists, entry/exit criteria, validation scripts per run |
Reading Order
Section titled “Reading Order”- Quick overview: This file →
decision-log.md(all decisions) →project-blueprint.md(dependency graph at the bottom) - Deep dive on a topic: Follow the links in the analysis index to the relevant document
- Implementation planning:
plan/choreography.md→ individualrun-N-*/project-plan.mdfiles - Implementation execution: Run-by-run, referencing analysis documents for design rationale; write outputs to
implementation/run-N-*/
Status
Section titled “Status”- Analysis phase: Complete — all design decisions recorded (45 decisions, see decision-log.md)
- Blueprint: Complete — 10 runs defined with verification gates
- Planning: Complete — detailed task plans for all 10 runs with validate-exit.sh scripts
- Implementation: Complete — all 10 runs executed, PR #59 merged to main (2026-03-20)
- Storybook polish: Complete — stories, MDX docs, ordering, Playground controls, cursor policy, sidebar molecule rendering
- CI: All checks green — build-and-test, VRT, CodeQL, Vercel
- External tickets:
Copyright: © Arda Systems 2025-2026, All rights reserved