refactor: separate workspace and transaction workflows.

Author: knightedcodemonkey

docs/pr-drawer-workspace-context-separation-plan.md

@@ -1,166 +1,46 @@
-# PR Drawer vs Workspace Contexts Separation Plan
+# PR Drawer vs Workspace Contexts Remaining Backlog
 
-## Purpose
+This document now tracks only work not yet implemented from the original separation plan.
 
-Define a post-4C UX split so transactional GitHub actions stay focused while local-first workspace lifecycle actions move to a dedicated surface.
+## Completed (for context)
 
-## Problem Statement
+1. Workspaces button and dedicated Workspaces drawer exist.
+2. PR drawer no longer exposes component/styles filename inputs.
+3. Commit targets are derived from workspace tab metadata.
+4. Checkbox copy uses entry-tab language.
+5. Confirmation summary includes a Files to commit list.
 
-The current Open Pull Request drawer mixes two distinct concerns:
+## Remaining Work
 
-1. Transactional sync actions:
-   - Open pull request
-   - Push commit to active pull request
-2. Workspace lifecycle management:
-   - Select/search local contexts
-   - Remove local contexts
+### Phase B follow-up
 
-This creates cognitive overhead and makes destructive local operations feel coupled to one-shot GitHub actions.
+1. Add multi-select removal in Workspaces drawer.
+2. Add richer filtering in Workspaces drawer.
+3. Decide whether to keep a quick context-switch affordance in PR drawer.
 
-## Goals
+### Phase C enhancements
 
-1. Keep PR drawer focused on transactional actions and immediate status.
-2. Move workspace lifecycle operations to a dedicated context management surface.
-3. Preserve local-first behavior and IndexedDB as source of truth for workspace content.
-4. Keep migration incremental and avoid breaking existing flows.
+1. Add open PR binding tools to Workspaces drawer.
+2. Add context health indicators in the Workspaces list (dirty, synced, drift).
+3. Add optional pin/favorite/recents support.
+4. Evaluate optional tab include/exclude toggles for commit targets.
 
-## Product Decisions (Locked)
+### Confirmation summary UX polish
 
-The following decisions are accepted for implementation:
+1. For long file lists, cap visible rows and show a +N more summary.
 
-1. Add a dedicated `Workspaces` control button in `app-grid-ai-controls`.
-2. `Workspaces` opens its own drawer for lifecycle operations (search/select/remove and future context features).
-3. Open PR / Push commit flows no longer expose `Component filename` and `Styles filename` fields.
-4. Commit target filenames are derived from workspace tab metadata stored in IndexedDB.
-5. The checkbox currently labeled `Include App wrapper in committed component source` will be relabeled to reflect entry-tab semantics.
-6. Open PR / Push commit confirmation summary shifts from two fixed file fields to a mapped tab/file list.
+### Modularization follow-up
 
-## Proposed Information Architecture
+Current implementation is primarily `src/modules/workspaces-drawer/drawer.js`.
 
-### 1. PR Drawer (Transactional)
+1. Split module if needed into smaller units:
+   - `state.js`
+   - `list-render.js`
+   - `actions.js`
+2. Keep PR transactional logic isolated in PR modules.
 
-Keep only action-scoped fields and status:
+## Validation Coverage to Keep
 
-1. Repository and branch selection
-2. File mapping summary derived from workspace tab metadata
-3. PR title/body and commit message
-4. Submit actions (Open PR, Push commit)
-5. Transaction status/errors
-
-Optional shortcut:
-
-1. Current context selector (read-focused quick switch)
-2. Link/button to open full context manager
-
-### 2. Workspace Context Manager (Lifecycle)
-
-Dedicated UI surface (modal or side panel) for:
-
-1. Search/filter local contexts
-2. Select/activate local context
-3. Remove one or many local contexts
-4. Future: open PR binding and context metadata management
-
-Location and trigger:
-
-1. Triggered from a new `Workspaces` button in `app-grid-ai-controls`.
-2. Implemented as a dedicated drawer separate from the PR drawer.
-3. Uses a colocated module structure under a parent directory dedicated to workspace lifecycle UI.
-
-## Storage Boundaries (Unchanged)
-
-1. IndexedDB:
-   - Workspace tabs and content
-   - Tab sync metadata (dirty/synced markers)
-   - Workspace context records
-2. localStorage:
-   - User preferences
-   - Lightweight per-repo PR drawer config
-
-## Migration Plan
-
-### Phase A: Transitional (Low Risk)
-
-1. Add `Workspaces` button in `app-grid-ai-controls` with dedicated icon.
-2. Add standalone `Workspaces` drawer skeleton and lifecycle list rendering.
-3. Move search/remove controls from PR drawer into `Workspaces` drawer.
-4. Keep quick context selection in PR drawer only as an optional shortcut.
-
-### Phase B: Consolidation
-
-1. Reduce PR drawer context UI to active context summary + switch shortcut.
-2. Remove component/styles filename fields from PR drawer.
-3. Derive commit file targets exclusively from workspace tab metadata in IndexedDB.
-4. Add multi-select removal and richer filters in manager.
-
-### Phase C: Follow-up Enhancements
-
-1. Open PR list binding tools live in manager.
-2. Context health indicators (dirty, synced, drift) appear in manager list.
-3. Optional pin/favorite/recents support.
-4. Optional tab-level include/exclude toggles for commit targets (if needed by workflow feedback).
-
-## Confirmation Summary Contract (Open PR / Push Commit)
-
-The confirmation dialog should show:
-
-1. Repository (`owner/repo`)
-2. Branch information:
-   - Open PR: `Base` and `Head`
-   - Push commit: `Head`
-3. Commit message
-4. A commit target list derived from tab metadata:
-   - Tab display name
-   - Repository-relative filepath
-   - Optional tag for entry tab
-
-Recommended rendering:
-
-1. Keep the top metadata lines concise.
-2. Show a bulleted list for `Files to commit` so users can quickly scan exact targets.
-3. For long lists, cap visible rows and show `+N more` summary.
-
-## Accessibility and UX Requirements
-
-1. Dedicated manager must support keyboard navigation for list/select/remove.
-2. Destructive actions must require explicit confirmation.
-3. PR drawer status remains transactional only.
-4. Context manager explains local-only deletion scope clearly.
-
-## Testing Plan
-
-1. PR drawer tests verify transactional workflows independent of context cleanup.
-2. Context manager tests verify search/select/delete workflows.
-3. Migration test verifies existing users retain contexts after UI split.
-
-## Open Decisions
-
-1. Modal vs side panel for context manager.
-2. Whether quick-select remains in PR drawer after Phase B.
-3. Whether context removal supports undo window.
-4. Whether the PR drawer should show all mapped tabs or only tabs marked as commit-included.
-
-## Implementation Structure Guidance
-
-To keep implementation modular and colocated:
-
-1. Create a parent module directory for lifecycle UI, for example:
-   - `src/modules/workspaces-drawer/`
-2. Keep small focused modules inside it, for example:
-   - `drawer.js` (controller)
-   - `state.js` (view state)
-   - `list-render.js` (UI rendering)
-   - `actions.js` (select/remove commands)
-3. Keep PR transactional logic in existing PR modules and consume workspace metadata via adapter functions only.
-
-## UI Copy Updates
-
-1. Rename checkbox label:
-   - From: `Include App wrapper in committed component source`
-   - To: `Include entry tab source in committed output`
-2. In summaries and status, refer to `entry tab` and `workspace files` rather than `component/styles files`.
-
-## Suggested Rollout
-
-1. Land after 4C stabilization tests are green.
-2. Ship Phase A first, then Phase B in follow-up PR.
+1. PR drawer tests remain focused on transactional workflows.
+2. Workspaces drawer tests cover search/select/delete and future multi-select behavior.
+3. Migration tests ensure existing stored contexts are retained.

docs/render-pipeline-multitab-spec-plan.md

@@ -1,136 +1,51 @@
-# Render Pipeline + Multi-Tab Spec Plan
+# Render Pipeline + Multi-Tab Vital Remaining Spec TODOs
 
-This document outlines test coverage to add after the render pipeline rewrite is fully integrated with the multi-tab UX.
+This file tracks only high-value gaps that still need coverage.
 
-## Why this plan exists
+## Already Covered (summary)
 
-Current test coverage intentionally removed a subset of specs that were tightly coupled to pre-rewrite assumptions:
+1. Entry-tab role behavior, rename stability, and restore behavior.
+2. Cross-tab import graph basics, missing-module and circular-import determinism.
+3. Core diagnostics/status flows, including pending/error/neutral transitions.
+4. Workspace persistence and per-repo context/config isolation baseline.
 
-1. Legacy assumptions around default-export hydration behavior in preview module assembly.
-2. Styles diagnostics behavior that depended on old compile/lint sequencing.
-3. PR drawer path validation timing assumptions tied to previous field sync flow.
+## Vital Remaining TODOs
 
-These should return as updated tests once the new pipeline contract is finalized.
+## 1. Default Export Support Matrix (High Risk)
 
-## Proposed Test Areas
+Add explicit render/typecheck specs for:
 
-## 1. Entry Resolution and Execution Semantics
+1. `export default class ...` in React mode.
+2. `function App() { ... } export default App` behavior.
+3. `const X = ...; export default X` behavior with entry-wrapper rules.
+4. Unsupported default-export combinations producing deterministic diagnostics.
 
-Goal: Validate how preview entry is resolved from workspace tabs under the `role: entry` model.
+## 2. Style Compile vs Lint Contract (High Risk)
 
-Add specs for:
+Lock down precedence and parity:
 
-1. Entry selection prefers explicit `role: entry`, with documented fallback behavior only when no explicit entry is present.
-2. Entry rename between `App.tsx` and `App.js` keeps execution stable.
-3. Entry path updates preserve directory while enforcing filename convention.
-4. Reload restores same entry tab and executes same source.
+1. Less error-path parity with Sass error behavior.
+2. Compile diagnostics + lint diagnostics precedence/coexistence contract.
+3. Clearing styles diagnostics does not affect component diagnostics/status.
 
-## 2. Default Export Handling in New Hydration Pipeline
+## 3. Status Aggregation Contract (High Risk)
 
-Goal: Reintroduce export-default tests against the final module assembly support matrix.
+Add state-machine coverage for:
 
-Add specs for:
+1. Multiple simultaneous error sources aggregating counts correctly.
+2. Clearing one scope updates only that scope and leaves other error states intact.
 
-1. `export default () => ...` in entry tab with manual render.
-2. `export default class ...` in React mode.
-3. `function App() { ... } export default App` compatibility.
-4. `const Button = ...; export default Button` behavior when App wrapper is implicit or explicit.
-5. Negative cases: unsupported default-export combinations produce deterministic diagnostics.
+## 4. Inactive Panel Mutation Guard (Medium Risk)
 
-## 3. Cross-Tab Import Graph Hydration
+Add keyboard/actionability spec that proves inactive editor panel input cannot mutate source.
 
-Goal: Ensure workspace graph resolution works across multiple component and style tabs.
+## 5. Update Obsolete PR Path-Validation Section (Doc/Test Hygiene)
 
-Add specs for:
+Old PR drawer filename-field validation cases are obsolete after tab-derived commit targets.
 
-1. Entry imports sibling component tab by relative specifier.
-2. Nested dependency chain (A imports B imports C) hydrates in stable order.
-3. Missing module path reports actionable preview error including unresolved specifier.
-4. Circular import emits stable error (or supported behavior) without hanging.
-5. Windows-style and POSIX-style separators normalize consistently in lookup keys.
+1. Replace with tab-derived commit target validation/normalization tests.
+2. Remove any remaining assumptions about component/styles filename fields in PR drawer flows.
 
-## 4. Styles Pipeline and Diagnostics Contract
+## Minimal Done Criteria
 
-Goal: Lock down expected diagnostics and status transitions for style dialects.
-
-Add specs for:
-
-1. Sass compilation error sets diagnostics state to error with styles-scope detail.
-2. Less error path behavior parity with Sass.
-3. Switching style mode clears stale diagnostics according to final pipeline contract.
-4. Styles lint diagnostics and compile diagnostics coexist or prioritize per contract.
-5. Clearing style diagnostics does not clear unrelated component diagnostics.
-
-## 5. Status and Diagnostics State Machine
-
-Goal: Ensure app status text/class and diagnostics toggle class remain consistent.
-
-Add specs for:
-
-1. Pending to error to neutral transitions for typecheck + lint + render.
-2. Multiple error sources aggregate counts correctly.
-3. Clearing one scope updates only corresponding status/diagnostics indicators.
-4. Auto-render off path keeps status stable until explicit render.
-
-## 6. Multi-Tab Tool Visibility and Actionability
-
-Goal: Guarantee controls are actionable only for active editor tab and panel.
-
-Add specs for:
-
-1. Component controls hidden/inert when styles tab is active.
-2. Styles controls hidden/inert when component tab is active.
-3. Keyboard interactions in inactive panel do not mutate source.
-4. Tab switches maintain tool visibility state and collapse state correctly.
-
-## 7. Persistence and Isolation Guarantees
-
-Goal: Verify deterministic startup and no stale state bleed between sessions.
-
-Add specs for:
-
-1. IndexedDB workspace restore across reload preserves tabs, active tab, entry role, and paths.
-2. PR drawer saved config does not unexpectedly overwrite active workspace tab paths.
-3. New session starts clean when storage is reset in tests.
-4. Repository switch behavior isolates per-repo local context and config.
-
-## 8. PR Drawer Path Validation and Sync
-
-Goal: Revisit path validation behavior after final field sync implementation.
-
-Add specs for:
-
-1. Reject traversal (`../`) for component and styles paths.
-2. Reject trailing slash paths for component and styles fields.
-3. Allow dotted segments that are not traversal.
-4. Entry-specific filename rule enforcement (`App.tsx` or `App.js`) reflected in drawer path values.
-
-## 9. Test Infrastructure Improvements
-
-Goal: Keep suites stable as UX evolves.
-
-Actions:
-
-1. Add helper APIs for tab activation before control interactions.
-2. Add one reset helper per suite to clear localStorage, sessionStorage, and IndexedDB.
-3. Prefer role/name selectors that match active-tab semantics.
-4. Avoid assertions that require hidden panel controls to be clickable.
-
-## Suggested Rollout Order
-
-1. Entry resolution + default-export support matrix.
-2. Cross-tab import graph hydration.
-3. Styles diagnostics contract.
-4. Status state machine.
-5. PR drawer path validation synchronization.
-6. Persistence/isolation hardening.
-
-## Definition of Done for this plan
-
-Before reintroducing removed specs, the render pipeline implementation should provide a written behavior contract for:
-
-1. Entry tab selection.
-2. Default export support matrix.
-3. Style compile + lint diagnostics precedence.
-4. Status/diagnostics state transitions.
-5. Path normalization and validation across workspace tabs and PR drawer fields.
+This plan is complete when the five sections above are covered by Playwright tests and linked from the affected suites.