Adds Agent Pro voting UI and tightens banner visibility logic in app/javascript/components/account-status-banner.js. Introduces hasAgentProSubscription prop, VISIBILITY_CHECK_INTERVAL, _visibilityTimer and renderKey state (used as PortalContainer key and in shouldComponentUpdate). Starts periodic DOM visibility checks and clears timers on unmount to avoid spurious re-renders and banner tampering; enables Agent Pro voting in production.

Introduce optimizeImage (lib/functions/image-optimizer.js) and tests (test/image-optimizer.test.js). The new action sends a note's embedded image to a Tinify shrink endpoint (defaults to a Plausible proxy), downloads the optimized bytes, replaces the note image in-place, and alerts the user with byte and percent savings. Tests call Tinify directly when an API key is present. Result: smaller images, reduced storage and bandwidth.

Update the Proposed Agenda plugin to propose schedule dates on the following day when the current day is too late. Adds proposed-agenda-next-day.test.js and moves relevant logic from the proposed-agenda service to cover this case. Impact: prevents invalid or missing today-time suggestions and provides sensible next-day proposals for users.

Adds minute-aware clock labeling (formatClockLabel(totalMinutes, timeFormat)) in lib/util/date-utility.js so times render minute-accurate AM/PM and 24h values (e.g. 570 → 9:30 AM). Exposes proposed-agenda in dashboard layout profiles and adjusts widget ordering/sizes to surface agendas. Increases shared-notes MAX_SCAN to 1000 in lib/shared-notes-service.js to broaden discovery — yielding finer time precision, improved agenda visibility, and better shared-note discovery.

Updates the shared-notes component (styles and tests updated) to support filtering notes by person and paginated results. Also annotates the person-note index with a build timestamp to record when the index was created. Impact: enables per-person note views, reduces payloads via pagination, and improves index management and troubleshooting.

Ensure the dashboard config popup is positioned inside the visible viewport. Adjusted popup positioning logic in lib/dashboard/dashboard-load.jsx and applyDashboardData in lib/dashboard/dashboard.jsx, added a small task-list tweak in lib/dashboard/dream-task.jsx, and updated styles (config-popup.scss, dashboard-layout-popup.scss). Result: popup no longer opens off-screen, improving accessibility and UX.

Add v2 LLM-driven Task Rescorer UI and agent (lib/functions/task-rescorer.js) using gpt-5.4-mini to propose clickable score recommendations with one-sentence rationales. Adds interactive task-source selection, rescore presets, configurable max-per-pass, concurrent batching, LLM quota/rationale limits, provider-selection logging and user API-key fallback; fixes by-tag lookup bugs. Improves reliability, observability and cost-control of rescoring and smooths the apply flow.

Adds a Layout Picker dashboard widget (lib/dashboard/layout-picker.js) with three one-click layout profiles and Apply/Dismiss plus persisted selection. Fixes Graveyard lazy-hydration by URI-encoding and persisting task->note metadata, switches to lazy getTask hydration, tightens logging and adds focused graveyard tests — resulting in faster retirement-candidate loads, preserved metadata, and predictable dashboard layouts.

Return humanEditedHours from parseNoteContent and append a human-edited footer in buildNoteBody so only user-edited Day Sketch hour lines are restored on load while auto-populated lines revalidate asynchronously. Remove brittle lib/util/new-component-popin.js, tighten recent-notes filtering and add logging to prevent schedule overwrites and flaky widget pop-ins — resulting in more reliable persisted Day Sketch edits and a stabler dashboard.

Introduce a v1 Graveyard dashboard widget by adding lib/dashboard/graveyard.js. This change adds the initial widget implementation (UI and supporting logic) to surface Graveyard items in the dashboard; it’s the first public iteration and is not yet deployed. Impact: enables the dashboard to display Graveyard content and provides a base for review and further enhancements.

Extract recent-notes behavior into lib/recent-notes-service.js and update the dashboard widget to persist per-day recent notes in an archived dashboard note (fenced JSON). Implement deterministic reseed (date+reseed hash), hydrate cached selections before render, and enforce discovery/scan/history caps. Update lib/dashboard/agenda.js to merge tasks and calendar events and sort by ascending time (all-day first; untimed last). Result: bounded dashboard state, stable reseeding, and correctly ord...

Add test/call-plugin-fallback.test.js which scaffolds and verifies calls to the Ample Agent plugin fallback used to provide AI features for the dashboard. The test builds mock app and fetch helpers and exercises the fallback invocation and response handling to validate the integration. Improves test coverage and reduces risk of regressions in dashboard AI calls.

Updated day-sketch to reliably strip HTML and Markdown when converting task content to plain text (lib/dashboard/day-sketch.js) by using a new stripMarkdown utility in lib/util/utility.js. Added Shift+Up/Down handling to extend selection on the current-hour input. Tests (test/day-sketch.test.js) and AI_CONTRIBUTIONS.md updated. Result: cleaner prefills and improved keyboard selection UX.

Refactor DaySketch by breaking a single large function into focused helpers and updating lib/dashboard/day-sketch.js to handle up/down arrow key navigation between entries. Adjusted prefill/use routines (entriesPrefilledFromTasks, prefillTextFromTask, useDaySketchEntries) for clearer flow. Result: improved keyboard navigation, clearer code structure and easier maintenance.

Fixed a bug in components/mood-rating-screen/post-rating-section.js where the mobile mood-rating response message contained the entire array of possible responses rather than one randomly selected entry. Updated the post-rating selection logic to pick and send a single response. This produces correct, concise messages and avoids sending oversized/incorrect payloads to consumers.

Introduce per-card preserve/complete/remove workflows in DreamTask with stable per-card keys, per-card dismissed/expanded state, CARD_REMOVE_FADE_MS timed removals, confetti on completion, and lifecycle callbacks (onPreserve, onComplete, onRemove, onToggleExplanation). Add a simple dev prompt modal (lib/util/browser-dev-app.js) that mocks app.prompt with checkbox/radio/select/date/text inputs to simplify developer testing and make card lifecycle behavior predictable.

Add a dashboard provider selection UI (lib/dashboard/dream-task-provider-selection.js) to let users choose which LLM to use when reseeding dream tasks if multiple models are available. Also update the default widget wrapper style to allow scrolling so the selector and long content fit constrained containers. Includes small docs and test adjustments — enables explicit model control and prevents layout overflow.

Surface the Replace Markdown action directly in the selection menu (app/javascript/lib/ample-editor/components/selection-menu.js): moved and consolidated the ConvertFromMarkdownButton logic and adjusted related ExtractContentButton code so replacing markdown no longer requires a second click when applicable. Deployed to production 2026-03-31 — reduces user clicks and streamlines editor workflow.

Detect markdown-like text in clipboard input and run it through the project's markdown parser (clipboard.js) instead of treating it as plain text HTML. Added logic to preserve blank lines as empty paragraphs and convert headings, bold and list markup into ProseMirror nodes, and added a between-apps test verifying OpenAI/Phind markdown is transformed correctly. Result: more faithful pastes from AI tools into the editor.