slate v2 synced blocks selection history coverage

Objective: Expand Slate v2 Synced Blocks selection/history coverage and list the current browser bugs around full-document keyboard navigation, Shift selection, Cmd/Meta+Arrow behavior, and history focus across repeated content-root projections.

Goal plan: docs/plans/2026-05-26-slate-v2-synced-blocks-selection-history-coverage.md

Template: docs/plans/templates/task.md

Primary template: docs/plans/templates/task.md

Applied packs:

• browser (docs/plans/templates/packs/browser.md)

Task source:

• type: user prompt
• id / link: local chat request
• title: Synced Blocks full selection/history coverage
• acceptance criteria: list the bugs; cover Arrow navigation through the whole document in both directions; cover Shift and Cmd/Meta+Arrow behavior; cover history focus/selection across synced copies; update /examples/synced-blocks fixture if more coverage needs another synced block from a different document.

Completion threshold:

• /examples/synced-blocks has enough fixture data to prove repeated synced copies and a different synced document root.
• Focused browser tests cover ArrowDown and ArrowUp through the logical document, active-copy history undo/redo, Shift+Arrow, and Cmd/Meta+Arrow rows.
• Safe root-owned bugs found by those rows are fixed in .tmp/slate-v2; larger bugs are listed with exact repro and owner.
• Targeted Synced Blocks browser tests and relevant slate-react package tests pass, or unresolved rows are intentionally marked as known failing/skipped with reasons.
• Task closure is legal only when the source-of-truth acceptance criteria are satisfied or explicitly narrowed, required verification evidence is recorded, code-review and release-artifact gates are closed when applicable, tracker/PR sync is complete or marked N/A with reason, and node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-26-slate-v2-synced-blocks-selection-history-coverage.md passes.

Verification surface:

• .tmp/slate-v2/playwright/integration/examples/synced-blocks.test.ts
• .tmp/slate-v2/packages/slate-react/test/content-root-navigation-contract.test.ts
• Targeted Playwright on http://localhost:3100/examples/synced-blocks
• Focused slate-react Vitest when runtime/navigation code changes.
• Completion checker for this plan.

Constraints:

• Preserve existing user-facing behavior outside the task scope.
• Prefer the durable ownership boundary over caller-by-caller patches.
• Do not create PRs, comments, commits, or pushes unless the task/user/skill requires them.
• Do not add broad ceremony when the task is trivial or docs-only.

Boundaries:

• Source of truth: user prompt in this turn plus existing Synced Blocks code.
• Allowed edit scope: .tmp/slate-v2/packages/slate-react/**, .tmp/slate-v2/site/examples/ts/synced-blocks.tsx, .tmp/slate-v2/playwright/integration/examples/synced-blocks.test.ts, and this plan.
• Browser surface: http://localhost:3100/examples/synced-blocks.
• Tracker sync: N/A, no tracker requested.
• Non-goals: Notion server sync, permissions, cross-page persistence, slate-yjs, and a broad selection engine rewrite beyond bugs proven here.

Blocked condition:

• Block only if the current selection model cannot represent a requested cross-root browser behavior without a larger architecture change; in that case record the exact failing row and owner instead of hiding it.

Task state:

• task_type: testing / browser regression coverage
• task_complexity: normal
• current_phase: closeout
• current_phase_status: complete
• next_phase: final response
• goal_status: ready_for_complete

Current verdict:

• verdict: complete for safe root-owned fixes and coverage expansion; one larger cross-root expanded-selection row remains listed with repro and owner.
• confidence: 0.86
• next owner: selection architecture follow-up for cross-root expanded ranges.
• reason: full-document ArrowUp/ArrowDown, ArrowLeft/ArrowRight repeated-copy traversal, Cmd/Meta+Arrow document boundary movement, root isolation, and history focus are covered by focused Playwright/package rows.

Completion rule:

• Do not call update_goal(status: complete) while any required checklist item remains unchecked. If an item does not apply, check it and add N/A: <reason>.
• Do not call update_goal(status: complete) until every completion threshold above is satisfied, final handoff evidence is recorded, and node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-26-slate-v2-synced-blocks-selection-history-coverage.md passes.
• Do not create hook state for this goal. This file plus the active goal are the durable state.

Start Gates:

Gate	Applies	Evidence
Skill analysis before edits	yes	task, testing, and autogoal loaded before edits.
Active goal checked or created	yes	Active goal created for this coverage slice.
Source of truth read before edits	yes	User prompt plus Synced Blocks source/test files read.
Tracker comments and attachments read	no	N/A: no tracker.
Video transcript evidence required	no	N/A: no video.
docs/solutions checked for non-trivial existing-code work	no	N/A: this is live browser coverage on new local feature surface.
TDD decision before behavior change or bug fix	yes	Add browser/unit regression rows before or with fixes.
Branch decision for code-changing task	yes	No PR/commit requested; keep current checkout as-is.
Release artifact decision	yes	Existing synced content-root changeset covers public slot/API; this slice adds tests/bugfix unless public API changes.
Browser tool decision for browser surface	yes	Use Playwright for committed browser rows and in-app Browser/browser-use for final route proof if code changes affect visible behavior.
PR expectation decision	yes	No PR requested.
Tracker sync expectation decision	yes	N/A: no tracker.
Browser pack selected	yes	Browser pack selected by scratchpad helper.
Browser route / app surface identified	yes	http://localhost:3100/examples/synced-blocks.
Browser tool decision recorded	yes	Playwright plus in-app Browser final proof if needed.
Console/network caveat policy recorded	yes	Runtime error guard for known Slate errors in browser tests; final browser proof records console caveat.

Work Checklist:

• Objective includes outcome, completion threshold, verification surface, constraints, boundaries, and blocked condition.
• Task source classified with source type, id/link, title, task type, acceptance criteria, caveats, likely files/routes/packages, browser surface, and root-cause layer.
• Required video or screen-recording evidence is cached/read as normalized <video-transcripts> XML, or marked N/A with reason.
• Nearby repo instructions and implementation patterns read before edits.
• Implementation fixes the right ownership boundary, or the narrower choice is recorded with reason.
• Release artifact requirement recorded: changeset, registry changelog, or N/A with reason.
• Final handoff shape decided: bug/feature/testing/batch/review/tracker requirements, PR body sync, and issue/Linear sync when applicable.
• Branch handling recorded for code-changing work: dedicated branch used, new branch needed, or N/A with reason.
• Local-env-rot retry policy recorded for any surprising repo-wide failure: reinstall/rerun evidence or N/A with reason.
• Workspace authority recorded: every proof command names the cwd/tool that owns the changed behavior.
• High-risk note recorded for public API, runtime, package-boundary, browser behavior, agent-action, or command-contract changes, or marked N/A with reason.
• Review/autoreview target selected from actual diff state for non-trivial implementation work, or marked N/A with reason.
• Agent-native review decision recorded for .agents/**, .claude/**, .codex/**, skills, hooks, commands, prompts, or user-action tooling.
• Browser pack: route, interaction path, and expected visible outcome are recorded before proof.
• Browser pack: browser proof uses the repo-approved browser tool or records a blocker/waiver.
• Browser pack: console and network errors are checked or explicitly out of scope.
• Browser pack: screenshot, trace, or exact verification caveat is ready for final handoff.

Completion Gates:

Gate	Applies	Required action	Evidence
Named verification threshold	yes	Done: run the named browser/package/typecheck/lint/check commands	bun check, focused Vitest, focused Playwright, all-project Synced Blocks Playwright, and completion checker close this threshold.
Bug reproduced before fix	yes	Done: encode rows in package/browser tests before closeout	Repeated-root exit/entry and history-focus bugs are covered by new regression rows; Shift+Arrow cross-root expansion is recorded as a larger follow-up.
Targeted behavior verification	yes	Done: run focused test/proof for changed behavior	Synced Blocks Playwright: 25 passed, 15 intentionally skipped outside Chromium; navigation contract Vitest passed.
TypeScript or typed config changed	yes	Done: run relevant typecheck	.tmp/slate-v2: bun --filter slate-react typecheck passed; bun check typecheck stage passed.
Package exports or file layout changed	no	N/A: no package export or file layout changes.	N/A: no pnpm brl required.
Package manifests, lockfile, or install graph changed	no	N/A: no manifest, lockfile, or install graph changes.	N/A: install not required.
Agent rules or skills changed	no	N/A: no .agents source edits.	N/A: generated skill sync not required.
Workspace authority proof	yes	Done: run verification in .tmp/slate-v2.	All package/browser commands were run with cwd /Users/zbeyens/git/plate-2/.tmp/slate-v2 except the root autogoal checker for this plan.
Browser surface changed	yes	Done: exercise route by Playwright browser rows.	PLAYWRIGHT_BASE_URL=http://localhost:3100 ... synced-blocks.test.ts passed on Chromium and all configured projects.
Browser final proof	yes	Done: use exact browser verification caveat.	Proof is automated Playwright route interaction coverage, not a manual screenshot.
CI-controlled template output changed	no	N/A: no template output changed.	N/A.
Package behavior or public API changed	yes	Done: add changeset.	.tmp/slate-v2/.changeset/synced-content-root-slots.md covers editable content-root slot and active-copy navigation/history behavior.
Registry-only component work changed	no	N/A: not registry-only component work.	N/A.
Docs or content changed	yes	Done: source-backed plan only.	This plan records evidence; no user docs changed.
High-risk mini gate	yes	Done: record failure mode and boundary.	Runtime/browser behavior risk is root-owner identity drift; fixed at runtime context/navigation boundary and proved by repeated-copy tests.
Agent-native review for agent/tooling changes	no	N/A: no agent/tooling changes.	N/A.
Local install corruption suspected	no	N/A: no failure matched local install corruption.	N/A: reinstall not run.
Autoreview for non-trivial implementation changes	yes	Attempted and blocked by tool/runtime failure.	Claude no-tools run hung and was killed; Codex no-tools rejected --no-tools; bounded Codex scoped run timed out after 180s with exit code 124 and no output.
PR create or update	no	N/A: no PR requested.	N/A: no check or PR body sync required.
PR proof image hosting	no	N/A: no PR body.	N/A.
Tracker sync-back	no	N/A: no tracker.	N/A.
Final handoff contract	yes	Done: fill fields below.	Final handoff records bugs, tests, caveats, and no PR/commit.
Final lint	yes	Done: run lint fixer and lint.	.tmp/slate-v2: bun lint:fix passed; bun lint passed.
Goal plan complete	yes	Done.	node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-26-slate-v2-synced-blocks-selection-history-coverage.md passed.
Browser interaction proof	yes	Done: exercise route interactions.	Full keyboard paths covered by Playwright on http://localhost:3100/examples/synced-blocks.
Browser console/network check	yes	Done by test failure policy.	Playwright route tests would fail on runtime errors; no separate network assertion was needed for local static example.
Browser final proof artifact	yes	Done: exact caveat recorded.	Automated Playwright proof only; no screenshot artifact.

Phase / pass table:

Phase	Status	Evidence	Next
Intake and source read	complete	Synced Blocks example/tests and runtime navigation/history paths inspected.	implementation complete
Implementation	complete	Runtime content-root owner registry, active-copy navigation repair, history focus repair, and expanded fixture/tests added.	verification complete
Verification	complete	Focused package/browser gates and bun check passed.	closeout complete
PR / tracker sync	complete	N/A: no PR, commit, push, or tracker requested.	final response
Closeout	complete	Plan evidence recorded; autogoal checker run before final.	final response

Findings:

• Existing Synced Blocks browser coverage does not exercise exiting the second mounted copy of a shared root. Source currently uses first-owner lookup for owner.childRoot === currentRoot, so repeated copies can exit at the wrong document position.
• Existing fixture has only one shared synced body root. Add a different synced document root so tests prove root isolation, not just repeated projection.
• Fixed bug: exiting a repeated shared content root used the first matching owner, so ArrowRight/ArrowDown/ArrowLeft/ArrowUp from the later copy could land beside the first copy instead of the active copy.
• Fixed bug: entering a later shared-root copy from the main document could focus the previously active or first mounted projection because target lookup was root-only instead of owner-path-aware.
• Fixed bug: keyboard undo/redo across main and synced roots restored content but could leave focus in the hotkey editor instead of the root whose operation was undone/redone.
• Fixed test bug: the synced body offset used a hard-coded 39; the current text length is 43, so tests now use the actual body string length.
• Larger bug left listed: Shift+ArrowDown across a root boundary creates a native newline selection while the Slate model selection stays collapsed in main and the synced root receives no selection. Owner: cross-root expanded selection / DOM-selection export architecture.

Decisions and tradeoffs:

• Put the fix in runtime content-root ownership and navigation/history repair, not in the Synced Blocks example. The example should be a consumer of the abstraction, not a pile of local focus patches.
• Add one separate synced document root to the example fixture. This is enough to prove repeated projection and root isolation without turning the example into a fake product.
• Do not force Shift+Arrow cross-root expansion into this slice. The model does not currently project a multi-root expanded range cleanly; hiding that under a local keyboard patch would be a bad fix.

Implementation notes:

• .tmp/slate-v2/packages/slate-react/src/hooks/use-slate-runtime.tsx and components/slate.tsx now expose runtime content-root owner registration, active-owner lookup, and owner-path-to-view-editor lookup.
• .tmp/slate-v2/packages/slate-react/src/components/editable-text-blocks.tsx registers each mounted content-root view against the owning element path.
• .tmp/slate-v2/packages/slate-react/src/editable/content-root-navigation.ts uses active owner state on root exit and owner-specific view lookup on root entry; it also covers Cmd/Meta+Arrow document boundary movement.
• .tmp/slate-v2/packages/slate-react/src/editable/keyboard-input-strategy.ts and runtime-keyboard-events.ts pass owner lookup into keyboard history repair.
• .tmp/slate-v2/packages/slate-react/src/hooks/use-slate-history.ts restores focus to the latest non-selection commit operation root before falling back to selection-root repair.
• .tmp/slate-v2/site/examples/ts/synced-blocks.tsx now renders two copies of one shared body root plus a separate synced document root.
• .tmp/slate-v2/playwright/integration/examples/synced-blocks.test.ts covers full ArrowDown and ArrowUp traversal, left/right repeated-copy traversal, Cmd/Meta+Arrow document boundaries, cross-root undo/redo focus, shared-root updates, separate-root isolation, duplicate, and unsync behavior.

Review fixes:

• Autoreview could not complete because the local review engines hung/timed out; no accepted autoreview findings were produced.

Error attempts:

Error / failed attempt	Count	Next different move	Resolution
Claude no-tools scoped autoreview hung silently	1	Kill and try Codex variant	Killed after roughly 2m47s.
Codex no-tools autoreview rejected --no-tools	1	Try bounded scoped Codex run without that flag	Immediate CLI rejection.
Bounded Codex scoped autoreview timed out	1	Record review blocker and rely on direct tests/checks	Timed out after 180s with exit code 124 and no output.

Verification evidence:

• .tmp/slate-v2: bun lint:fix passed.
• .tmp/slate-v2: bun lint passed.
• .tmp/slate-v2: bun --filter slate-react typecheck passed.
• .tmp/slate-v2/packages/slate-react: bun test:vitest -- test/use-slate-history.test.tsx test/content-root-navigation-contract.test.ts passed: 2 files, 15 tests.
• .tmp/slate-v2: PLAYWRIGHT_BASE_URL=http://localhost:3100 PLAYWRIGHT_RETRIES=0 bunx playwright test playwright/integration/examples/synced-blocks.test.ts --project=chromium passed: 10 tests.
• .tmp/slate-v2: PLAYWRIGHT_BASE_URL=http://localhost:3100 PLAYWRIGHT_RETRIES=0 bunx playwright test playwright/integration/examples/synced-blocks.test.ts passed: 25 passed, 15 skipped. Skips are intentional non-Chromium skips for geometry/focus rows.
• .tmp/slate-v2: PLAYWRIGHT_BASE_URL=http://localhost:3100 PLAYWRIGHT_RETRIES=0 bunx playwright test playwright/integration/examples/editable-voids.test.ts --project=chromium --grep "content root|vertically|clicking outside" passed: 3 tests.
• .tmp/slate-v2: bun check passed: lint, package/site/root typecheck, package tests, slate-layout, and slate-react Vitest.
• /Users/zbeyens/git/plate-2: node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-26-slate-v2-synced-blocks-selection-history-coverage.md passed.
• Cleanup scan: rg -n "console\\.log|test\\.only|\\.only\\(|debugger" ... found only the pre-existing android input manager debug comment.
• Shift+Arrow probe: from main p1 at [0,0] offset 2, Shift+ArrowDown produced native selected text "\n", kept outer Slate selection collapsed at [0,0] offset 2, and left the first synced root selection null.

Final handoff contract:

• PR line: no PR, commit, or push requested.
• Issue / tracker line: no tracker requested.
• Confidence line: high for navigation/history fixes and coverage; medium for the broader Shift+Arrow follow-up because it needs explicit multi-root range model design.
• Flow table:
  • Reproduced: repeated-root exit/entry and history focus rows are reproduced by package/browser tests; Shift+Arrow bug reproduced by one-off browser probe.
  • Verified: focused package tests, Synced Blocks Playwright, editable-voids regression subset, lint, typecheck, and bun check passed.
• Browser check: automated Playwright proof on http://localhost:3100/examples/synced-blocks.
• Outcome: example fixture and coverage now exercise repeated shared roots, a separate synced document root, full keyboard traversal, Cmd/Meta+Arrow, and cross-root history focus.
• Caveat: cross-root expanded Shift selection remains a real bug and is not papered over.
• Design:
  • Chosen boundary: runtime-owned content-root owner/view registry plus navigation/history repair.
  • Why not quick patch: example-local keyboard code would still break the same abstraction for the next synced/rooted element.
  • Why not broader change: full multi-root expanded selection is a larger selection architecture task; this slice fixes safe owner identity bugs and records the remaining row.
• Verified: see verification evidence.

Final handoff / sync:

• PR: N/A, no PR requested.
• Issue / tracker: N/A, no tracker requested.
• Browser proof: Playwright interaction proof on the Synced Blocks route; no screenshot artifact.
• Caveats: autoreview tool blocked; Shift+Arrow cross-root expansion remains.

Timeline:

• 2026-05-26T10:47:23.099Z Task goal plan created.
• 2026-05-26T12:50:00Z Added repeated shared-root and separate synced-document fixture coverage.
• 2026-05-26T13:25:00Z Fixed active owner lookup for repeated content-root projections.
• 2026-05-26T13:45:00Z Fixed history focus repair across main/synced roots.
• 2026-05-26T14:20:00Z Ran focused and full verification gates.
• 2026-05-26T14:45:00Z Recorded remaining Shift+Arrow cross-root selection bug and autoreview blocker.
• 2026-05-26T14:50:00Z Autogoal completion checker passed.

Reboot status:

Question	Answer
Where am I?	Closeout complete; waiting for final response.
Where am I going?	Run the mechanical autogoal checker, mark the active goal complete, then hand off.
What is the goal?	Synced Blocks selection/history/browser coverage with bugs listed and safe owner bugs fixed.
What have I learned?	Repeated content-root projection needs active owner identity; Shift+Arrow across roots needs a larger expanded-selection design.
What have I done?	Added fixture coverage, package tests, Playwright coverage, runtime/navigation/history fixes, and verification evidence.

Open risks:

• Cross-root expanded Shift selection is still broken by design gap, not a local Synced Blocks fixture bug.
• Autoreview did not finish because local review engine runs hung or timed out.