--- title: "Release notes" description: "Coassemble v2.0.1" canonical_url: "https://support.coassemble.com/articles/release-notes-HcvM2zzVCE" md_url: "https://support.coassemble.com/articles/release-notes-HcvM2zzVCE.md" --- # **Coassemble v2.0.1** ## **New Features** ### **Screen embeddables** Embed customers can now work with a single screen rather than a whole course. Two new embeddable types: **Screen Builder**, which puts the Builder 2 screen editor inside your own application, and **Screen Player**, which renders one screen for a learner. * **Start from a layout** — adding a screen opens the layout picker, so authors choose a design and begin with placeholder content already in place * **Build your own picker instead** — list the available layouts and create a screen directly, if you'd rather design the experience yourself * **Quizzes included** — a standalone quiz embeds as a set, with scoring, retakes and results * **Your product stays in control** — screen start, completion, answers and scores are sent to the host application, and the built-in navigation can be hidden so your own controls take over * **Optional AI assistant** — off by default, and scoped to editing the screen it's on ### **SCORM hosting** Embed customers can upload an existing SCORM package and host it in Coassemble alongside authored courses. * **Sits in the library like any other course** — packages appear as course cards with an editable cover image and a SCORM label * **Attempt tracking** — see attempts, time spent, success and score, compare multiple attempts side by side, and grant an extra attempt to a learner who is genuinely blocked * **Question-level export** — download individual question results as a CSV you can pivot * **Clear status while processing** — an explicit state while a package extracts ## **Builder 2 improvements** ### **Mobile layout editing** Switch the editor to mobile and control how a screen stacks on a phone, separately from the desktop design. * **Hide blocks on mobile** — leave out anything that doesn't earn its space on a small screen * **Reorder for mobile** — set a different block order without touching the desktop layout * **Sensible image heights** — images keep the shape you cropped rather than stretching to full phone width, and can be resized from the bottom edge ### **Reposition images** Choose which part of an image shows inside a block, instead of only fitting or filling it. * **Crop, pan and zoom** — drag to reposition, zoom up to 3×, with the rest of the image shown faded so you can see what's outside the frame * **Works inside cards too** — available for images in process, flashcards and slideshow blocks * **Existing courses are untouched** — images already placed look exactly as they did ### **Custom fonts on headings and paragraphs** Pick a font outside your theme's set directly from the heading and paragraph toolbars, with weights grouped into their own menu and a one-click reset back to the theme font. ### **Better AI images** * **Images now reflect your course topic**, not just the screen they sit on * **Generation shows its progress** — the builder marks which screen is being illustrated and follows along ### **Bug Fixes** * Fixed **menus** drifting out of position when scrolling, and menus that scrolled when they didn't need to * Fixed **insights** showing empty cards that looked like they were still loading — the selected period is now named * Fixed **narration** continuing to play on screens with no narration recorded * Fixed **focus rings** appearing on Classic course titles when a screen loads * Improved **stability** of the course player under load # **Coassemble v 2.0** ## **New Features** ### **Builder 2 — now in public beta** A new way to create courses, built around a conversation. Bring what you have, answer a few questions, and the agent builds a structured course with you. * **Start from almost anything** — Word, PowerPoint, PDF and Keynote documents, or just a written brief * **Choose how closely to follow your source** — Verbatim keeps your wording as written for policies and compliance, Structured reorganises it into a clear learning flow, Adaptive rewrites and expands it into a fuller course * **A conversation that adapts** — it asks only the questions that matter, skips anything it can work out from your material, and checks your answers make sense before building * **Courses group into sections** — learners see them as chapters, each with its own progress ### **A new way to build screens** Every screen is made of blocks you can arrange, swap and restyle — with layouts that keep courses looking designed rather than assembled. * **Content blocks** — headings, text, images, video, audio, tables, buttons and embedded documents * **Interactive blocks** — accordions, checklists, process steps, flashcards and slideshows * **Assessment blocks** — multiple choice, true/false, dropdown and matching questions, plus scored quizzes with results * **Layouts do the design work** — screens follow proven templates instead of freeform placement, so spacing and hierarchy stay consistent * **Control the reading order** — set the order blocks are read by screen readers, narration and keyboard navigation ### **Refine your course by asking** The agent stays with you in the builder. Describe what you want changed and it does it, rather than you hunting through menus. * **Generate and rewrite** — add screens, rework wording, adjust tone or reading level * **Add assessment** — insert knowledge checks or a quiz where you need one * **Restructure** — reorder screens, split sections, reshape the course * **Beyond content** — narration, translations, theming, branding and images are all in reach * *Available on selected plans* ### **Improve courses based on how learners actually did** The agent can read your course analytics, so you can ask what's working and act on the answer without switching context. * **Find the drop-off** — screen-by-screen completion shows exactly where learners stop * **Ask in plain language** — "where are people giving up?" or "which questions are people failing?" * **Fix it there and then** — the same conversation that spots a weak screen can rewrite it * **Narrow it down** — filter by date range to compare a recent cohort against an earlier one * *Available on selected plans* ### **Build in your language** The course builder is now fully localised, so your team can work in the language they think in. * **Eleven languages** — English, French, German, Spanish, Italian, Portuguese, Dutch, Danish, Swedish, Norwegian and Finnish * **The agent follows suit** — write to it in French and the whole conversation happens in French, including questions, options and the final summary * **Independent of course language** — plan a Spanish course in English if that's easier for you ### **A refined learner experience** Builder 2 courses have a rebuilt learner view, designed to keep people oriented and moving. * **One button drives progression** — no hunting for what to do next * **Progress you can see** — a bar across the top shows position in the course and progress through the current screen * **A course outline sidebar** — sections, screens and what's been completed, at a glance * **Pick up where you left off** — returning learners are offered a resume prompt ### **Bring your existing SCORM courses** Upload a SCORM package and convert it into a fully editable Coassemble course. * **Supports the major authoring tools** — Articulate Rise and Storyline, Adobe Captivate and iSpring * **Keeps your content** — text, images, galleries, flashcards, process steps, accordions and knowledge checks carry across * **Tells you what didn't convert** — anything that couldn't be brought over is listed for review rather than silently dropped ### **Course Coach** Built-in expert feedback on your course, on demand. * **Reviews instructional quality, engagement and accessibility** — with specific, actionable findings rather than a score * **Points straight at the problem** — each finding links to the screen it refers to * *Available on selected plans* ### **Connect Coassemble to your AI assistant** A new MCP server lets assistants such as Claude create and manage courses directly in your workspace. * **Over 50 tools** covering course creation, editing, publishing and analytics * **One sign-in** — connect with a single OAuth authorisation, managed from Settings ### **Drive the agent from your own product** If you embed Coassemble, your application can now drive the same agent over MCP or REST. * **Scoped and controlled** — limit it to a single course, choose which tools are available, and set it read-only if you need to * **The same capabilities** — your users get the agent experience without leaving your product ## **Bug Fixes & Improvements** * Replaced **Unsplash** with **Pexels** as the built-in image library * Fixed **quiz retry limits** not applying after back navigation, and retries not clearing tile progress * Fixed **unscored quizzes** blocking navigation, completing early, or emitting completion twice * Fixed an issue where **quiz completion** wasn't recalculated after dynamic questions were added * Fixed **quiz feedback** showing an empty image placeholder * Fixed **video uploads in the EU and UK** downloading the source file instead of processing it * Fixed **uploaded videos** and **embedded PDFs** missing from standalone SCORM exports * **Translations** are now included in static SCORM exports * Fixed **narration** missing bullet points and checklist sub-items when generating audio * Fixed **narration tracks** not swapping when the default course language changed * Fixed **course language** being lost when a published snapshot omitted it * Fixed the **insights date filter** excluding recent completions, and tracking detail pages returning errors * Fixed **quiz scores** not being returned for courses delivered through the API * Fixed **DOCX previews** getting stuck on "Generating preview" * Improved **PDF handling** — retries on network failures, and support for older browsers * Fixed **API keys** being regenerated every time the API settings were opened * Fixed **embed URLs** and the embed preview pane not fitting the viewport * Fixed **Microsoft Teams** re-asking for consent on already-authorised tenants * Fixed **Slack** access links for Flow plan courses * Improved **course thumbnails** — fonts and images now finish loading before capture * Security hardening on **outbound URL handling** # **Coassemble v1.8.3** 26th May 2026 v1.8.3 launches **Embeddables** — a new way to drop Coassemble learning experiences, authoring tools and analytics dashboards directly into your own apps and platforms using signed-URL iframes, backed by workspace-level API keys and a new developer documentation site. Alongside Embeddables, image generation moves out of beta on a new model with a manual trigger, AI narration adds two new voice locales, and a wave of authoring, video and translation improvements ship. The release also wraps in the hotfixes that have gone to production since v1.8.2. ## New Features ### Embeddables Coassemble Embeddables go live in v1.8.3 as a fully supported, customer-facing capability. Customers can drop Coassemble learning experiences, authoring tools and analytics dashboards directly into their own apps and platforms using signed-URL iframes. A new developer documentation site walks through every embeddable component end to end, with live code examples. Availability of individual embeddables varies by plan — see the developer docs for the current plan matrix. A standout in this release: **Collections are now first-class in Headless for the first time.** Until now, Headless only supported courses. v1.8.3 introduces a full Collection REST API (GET /collections, GET /collections/{id}, POST /collection, DELETE /collection/{id}, POST /collection/{id}/restore) and a full set of Collection embeddables — Player, Builder, List, Card and Analytics — so integrations can programmatically create, manage and embed collections of courses end to end. * **Course Player and Collection Player** — embed interactive courses and collections for learners, with per-user signed URLs * **Course Card and Collection Card** — embed a single course or collection as a clickable preview card * **Course List and Collection List** — embed a browsable list of courses or collections as card grids * **Course, Collection and User Analytics** — embed read-only dashboards showing engagement, completion and user-level activity * **Course Builder and Collection Builder** — embed Coassemble's authoring and collection-management tools directly inside your product * **course.generated postMessage event** — the embedded Course Builder now posts a course.generated event to the parent window when AI course generation completes, with the course's id, title and screenCount, so host apps can react the moment a course is ready * **Workspace-scoped API keys** — a new COASSEMBLE:\:\ key format replaces the old user-level keys, so integrations no longer break when the owning user rotates a key or leaves the workspace. Existing COASSEMBLE-V1-SHA256 keys continue to work * **Tiered API actions** — API keys now carry capability tiers, so each plan only unlocks the embeddables and endpoints included in its tier * **New developer documentation site** — full reference for embeddables, the REST API (Courses, Collections, Themes, Identities, Tracking, Translations, Usage, Generate), webhooks, events and errors ### **Image generation: now manual, powered by Gemini Flash 2.5** AI image generation is moving out of beta with an upgraded model — Gemini Flash 2.5, replacing Imagen (which Google is sunsetting). Image generation is now triggered on demand instead of running automatically, giving authors full control over when images are created and tokens are used. * **Manual generation** — images are no longer auto-generated when creating courses from a prompt, transforming documents, or generating screens * **Monthly token allowance** — a monthly image generation allowance is now in effect on some plans, with a clear usage indicator in the AI image modal. Existing headless customers are not affected ### **New narration voices** * Added **French (Canada)** and **Spanish (Mexico)** voices to AI narration, expanding language coverage for global teams ### **More reliable narration** A round of improvements to make narration generation faster to recover from transient issues and clearer when something goes wrong. * **Categorised publish errors** — the publish modal now groups narration errors with targeted retry and edit actions, so authors know exactly what to do next * **Smarter retries** — narration generation now retries on transient Google TTS timeouts and 503s, with a long-text fallback for big scripts * **Automatic re-generation triggers** — adding, removing or altering questions now correctly flags narrations for regeneration * **Cleaner script handling** — auto narration is preserved when a custom script is absent, and the narration modal's language is now read-only as intended ## **Bug Fixes & Improvements** * Fixed an issue where **video playback** would fail to recover when Mux returned a 404 or had no static rendition available — playback now falls back gracefully and retries on 429 responses * Resolved **YouTube embeds** failing to initialise with the enablejsapi flag, and **YouTube fullscreen** breaking inside nested SCORM iframes * Fixed **Vimeo oEmbed** producing a duplicate url parameter on the XHR request * Resolved a **video status polling** issue that was being incorrectly caught by the global rate limit * Fixed **course player crashes** where the screens array could be missing or malformed — added guards across navigation, completion and oEmbed restoration * Resolved a **course revert** issue where the course theme wasn't returned, so revert sometimes left themes out of sync * Fixed an issue where **quiz generation** would occasionally not return any questions when creating a new screen * Fixed **finish screen narration** properties not persisting, and **paragraph breaks** being stripped from narration text * Resolved **standalone narration** not playing when narration was not enforced or there was no start button * Fixed a **folder restoration** bug where restoring a course would clear its folderId instead of returning it to its original folder * Resolved an issue where **adding to a folder** before it had loaded would silently fail * Fixed **paste into the email-multi share field** being broken after a recent text input change * Resolved an issue where **reordering questions** unnecessarily bumped the course's updated timestamp # **Coassemble 1.8.2** 31 March 2026 ## **New Features** ### **Improved Folder Functionality** We've made it easier to discover and use folders on your dashboard, plus added new ways to organise your courses. * A new **Folder** button in the dashboard header makes it clearer how to create folders * **Drag and drop** courses into existing folders to organise them quickly * **Drag a course onto another course** to automatically create a new folder containing both * Available on plans with folder access ### **Sign in with Slack** You can now use your Slack account to sign up and sign in to Coassemble. * **Sign in with Slack** is available on the sign-in screen, registration flow, and when returning with an existing account * One-click authentication — no extra password to remember * Works alongside existing sign-in options like Google and email ### **Bug Fixes & Improvements** * Fixed an issue where **quiz generation** could fail silently — errors are now handled gracefully * Resolved **video playback** issues related to the Mux migration, including missing static renditions and metadata formatting * Fixed a bug where **course narration** wouldn't play correctly when the start button was hidden * Fixed **learner progress tracking** failing in certain edge cases * Resolved an issue where **reverting a course** didn't properly restore the course theme * Fixed **quiz screen creation** so questions are always returned correctly when adding new screens * Improved **video upload** reliability, including better retry handling when the video service is busy # **Coassemble 1.8.1** 17 March 2026 ## **Major Features** ### **Brand Voice** Introducing **Brand Voice**, a new way to ensure AI-generated content matches your organisation's tone and style. You can now: * Define your brand's **primary voice** and adjust traits like friendly vs. professional, casual vs. formal, and more * Set a **default language** for all AI-generated content * Add **writing examples** so the AI learns your specific style and rhythm * Specify preferred **vocabulary** — terms to favour and terms to avoid * Choose an **image generation style** (Photo, Cinematic, Drawing, Vector, Painting) and exclude unwanted imagery Available on Enterprise plans. ### **Video Transcoding Infrastructure** We've replaced Mux with a new region-locked video transcoding pipeline built on **AWS MediaConvert**. While Mux supported region-locking for US and EU, it did not support the UK. * All video uploads, transcoding, and delivery now stay entirely within their deployment region (US, EU, or UK) * UK enterprise customers can now meet the same **data residency** requirements previously only available in US and EU * Videos are transcoded into multiple quality levels with **adaptive bitrate streaming** * No changes to your day-to-day workflow ### **Webhooks** Subscribe to real-time event notifications for your courses via webhook endpoints. * Configure endpoints in **Settings → Webhooks** to receive POST requests for key course events * Each delivery includes course details, learner information, and a **signing secret** for verification * Failed deliveries are automatically retried ### **Video Asset Library** Videos can now be uploaded and managed in your **Asset Library** alongside images. * Browse, search, and reuse video assets across courses without re-uploading * **Multi-file uploads** are supported * Videos display real previews in the library grid ### **Translations API** A new public API for programmatically managing course translations, ideal for headless integrations. * Create, sync, and manage translations via API endpoints * Set default languages and sync after source content updates ## **Improvements** * Collection admins can now customise **finish screen navigation** when courses are part of a collection, configuring different button labels and destinations for mid-collection vs. final courses. * On screens with unsynced narrations, a single standalone **voiceover button** now appears instead of individual item-level narrations. * When advancing through quiz questions, focus now automatically moves to the new question content, improving the experience for **keyboard and screen reader** users. * Internal infrastructure for Sevalla-hosted environments (EU and UK) is now codified with **Terraform**. * Shared template links now work for people who don't yet have a Coassemble account. ## **Bug Fixes** * Pop quiz results now correctly show a failing message when only a few answers are provided, instead of incorrectly reporting a pass. * Deleted quiz questions are no longer included in tracking API responses. * **Collection themes** are now correctly applied when viewing a collection via a direct link in the app. * Fixed checkbox selection not responding to clicks. * Exporting a filtered **Activity Report** now exports only the filtered data, not the entire dataset. * **Brand Kit** logos now display correctly when a custom theme is applied. * Autoplay narrations no longer play on the course preview screen when selecting a microcourse module. # **Coassemble 1.8.0** Feb 23, 2026 ## **Major Features** ### **Brand Kit** Introducing **Brand Kit**, a new way to manage and apply your organisation’s branding across Coassemble. You can now: * Create and save reusable **custom themes** (colours, typography, body text styling) * Upload and manage multiple **logo variants** (SVG, PNG, JPG) * Access your **Image Library** from within Brand Kit * Upload custom fonts (WOFF, WOFF2, TTF, OTF) ### **Course-Level Quiz Success Status (SCORM)** Quiz results now roll up to the course level and emit a success status via SCORM. * Single-quiz courses report the quiz result * Multi-quiz courses calculate success using the average score across all quizzes * Supported for Standalone and Dynamic SCORM ## **Improvements** * Users are now taken directly into the course editor immediately after course generation completes. * Introduced a 50MB file size limit for document transformation and conversion processes to improve reliability. * Added standalone **Share** and **Insights** buttons to Collection cards. * Summary statistics now load immediately without requiring all learner data to be manually loaded. * Implemented final updates following the second round of independent audit for WCAG accessibility assessment. ## **Bug Fixes** * Collection Insights summary statistics now display correctly without needing to wait for all learner data to load. * Fixed an issue where sharing to a Slack channel did not generate a pre-authorised link. * Changing a course theme before taking any other action will no longer prevent you from adding screens. * The **Convert a Presentation** flow now correctly prompts you to import images. * Narration scripts will no longer include the name of the video file. *** # **Coassemble 1.7.9** **14 January 2026** ## **Improvements** ### **Accessibility & Compliance** * **WCAG 2 Compliance:** Implemented several accessibility enhancements following an independent audit to ensure a more inclusive experience for all users. ### **API & Headless** * **AI Control for Headless:** Added a new API flag that allows Headless customers to completely disable all AI functionality. This supports organizations with strict policies regarding AI-driven services. * **Revert Endpoint Stability:** Resolved a technical issue where calling the revert endpoint caused the clientIdentifier to return as null. ### **Content & Learning Experience** * **SCORM Progress Persistence:** Implemented a "progress restore state" for Static SCORM packages, ensuring learners can pick up exactly where they left off. ## **Bug fixes** * Fixed a bug where quiz questions were appearing in a random order rather than the designated sequence. * Corrected an issue where the wrong quiz attempt data was being displayed in the Insights tab. * Resolved an issue that allowed multiple sign-ups using the same email address. * Fixed a Firefox-specific bug that prevented long titles from saving on the Start screen. * Patched an issue that prevented some users from progressing past the account verification step. *** # **Coassemble 1.7.8** **10 December 2025** ## **New Features & Improvements** ### **Narrations for translations** Users can now generate narrations for translated courses. The voice is automatically chosen depending on the language of the translation. ### **SCORM Tracking records available via API** SCORM tracking data is now accessible through the API, enabling Headless customers to view and manage records programmatically. This also allows use of the existing delete endpoint: `DELETE /api/v1/headless/tracking.` ### **Duplicate courses across workspaces** Creators with multiple workspaces can now duplicate a course from one workspace to another, simplifying multi-workspace workflows and content distribution. ### **Template Links for course sharing** Courses can now be shared as a template that can be edited in another workspace, even those they are not a member of. ### **Standalone SCORM Progress Events** Standalone SCORM packages now emit progress events on every screen, ensuring more accurate tracking and smoother integrations. *** # **Coassemble 1.7.7** **17 November 2025** ## Major Features ### Detailed quiz insights for learners and creators Learners and creators now have access to a detailed breakdown of quiz results. Learners can review their performance immediately after completing a quiz. Creators can view detailed attempt data in Course Insights. ### Improved image generation Coassemble now uses Google's Imagen for AI-generated images. ### Learning outcomes step for AI course creation The AI course creation flow now asks for learning outcomes before generating your course. ## Improvements * You can now set custom thumbnails for lessons. * Navigation and pagination refinements. * Added Slack user verification. * Images now preload on course load. ## Changes to data processing and AI Improved AI processing performance and reliability. Connected directly to Gemini Flash and Imagen. Convert API now region-locked to US. ## Bug Fixes * Fix for re-listening narration issue. * Image generation failure fix. * Undo/Redo fix. * Standalone SCORM compile fix. * Quiz reattempt settings fix. *** # Coassemble 1.7.6 **5 November 2025** ## Major Features ### AI Narrations You can now add high-quality AI voiceovers to courses. ### Course Publishing Control when updates go live by publishing drafts. ## Minor Improvements * Consolidated thumbnail generation. * Enhanced presentation conversion. * Flashcards alignment updates. * Accessibility fixes. * Mobile navigation improvements. ## Bug Fixes * Answer add button fix. * Hotspot empty title fix. * True/False auto-correct selection fix. *** # Coassemble 1.7.5 **29 September 2025** ## New Features & Improvements ### Loom Integration Users can record their screen directly. ### Alternative text management Improved handling and reuse of alt text. ### Flashcards update Added flip button for learners. ## Bug Fixes * Match Corresponding response fixes. * Quiz question visibility fix. * Updated email preferences fix. * Export performance improvements. * SCORM completion fix. * Collection insights accuracy fix. * Duplicate email creation fix. *** # Coassemble 1.7.4 **22 September 2025** ## New Features & Improvements ### POST route for signed URLs New POST route added; GET deprecated. ### Headless usage report enhancement Added total course views. ### Loader for dynamic SCORM files Improved load experience. ## Bug Fixes * Static SCORM completion fix. * Incorrect collection insight fix. * Collection insights load fix. *** # Coassemble 1.7.3 **25 August 2025** ## New Features & Improvements ### Slack Integration Share courses directly to Slack; real-time notifications. ### Headless Creator Reports Detailed creator activity reports. ### Video Backup Service Videos backed up to AWS S3. ## Fixes & Improvements * ALT tag saving fix. * Document image import fix. * Collections layout fix. * Accordion item removal fix. *** # Coassemble 1.7.2 **04 August 2025** ## New Features & Improvements ### Redesigned Share Modal Streamlined UI with Approved Domains. ### Approved Domains Restrict access by domain. ### New API Endpoint Programmatic course generation. ## Other Improvements * Better undo/redo tracking. * Quiz scores added to Insights exports. * Vimeo playback speed defaults enabled. ## Bug Fixes * Match Corresponding visibility fix. *** # Coassemble 1.7.1 **22 July 2025** ## New Features & Improvements * Multiple Document Support. * Increased max course duration. * Improved quiz defaults. * Better progress tracking. * Tag support for Headless. * Error notifications. * Back navigation for AI creation. ## Bug Fixes * Share modal removal issue. * Collections not saving. * Completed panel empty fix. * Google Drive Headless fix. # Coassemble 1.7.0 **14 July 2025** ## New Features & Improvements ### Standalone SCORM export Allows export without Coassemble connection. ### New Enterprise plan Includes Translations and Standalone SCORM. ## Miscellaneous * Improved Google Drive integration. * Added timestamps to exported reports. # Coassemble 1.6.10 **23 June 2025** ## New Features & Improvements * Google Drive integration. * Microsoft OneDrive integration. * Headless API updates. ## Bug Fixes * Fixed incorrect email linking. # Coassemble 1.6.9 **23 June 2025** ## New Features & Improvements * Translations for Headless. * Quiz Answer Storage API access. ## Bug Fixes * Share modal input fix. * Duplicate quiz deletion fix. * Quiz save issues fixed. * Hotspot image replacement fix. # Coassemble 1.6.8 **02 June 2025** ## New Features * Presentation-to-course conversion. * API support for conversion. * Editable Flashcards finish state. * True/False translation support. * Finish button URL setting. * Extend Image feature. ## Bug Fixes * Video resume fix. * Thumbnail autoplay fix. * Flashcard stack fix. * Mobile hotspot modal fix. # Coassemble 1.6.7 **19 May 2025** ## Onboarding Experience New flows letting users preview AI course builder before signup. ## Bug Fixes & Improvements * Improved image relevance. * Flashcards display fix. * Hotspot mobile image fix. # Coassemble 1.6.6 **29 April 2025** ## New Features * New AI assistant for course creation. ## Improvements * Start/Finish email notifications. * Dashboard shortcuts. * New Headless endpoints. # Coassemble 1.6.5 **31 March 2025** ## New Features * AI-powered Screen Generation. * SCORM Anywhere sharing. * Headless enhancements. ## Improvements * Hotspot reordering. * AI-generated quizzes. * Image generation improvements. * Insights infinite scroll. # Coassemble 1.6.4 **20 March 2025** ## New Features * Refine feature. * AI image generation (BETA). * API theming support. ## Improvements * Sidebar redesign. * Removed automatic Unsplash images. * Rich text for checklist. ## Bug Fixes * Collections image alignment fix. # Coassemble 1.6.3 **13 March 2025** ## New Features * Workspace ownership transfer. * Insights export. * Duplicate endpoint. ## Improvements * Insights performance. * Match Corresponding updates. * Thumbnail generator rebuild. * Saved dashboard sort. ## Bug Fixes * Document parsing fix. # Coassemble 1.6.2 **11 March 2025** ## New Feature: Collections Group multiple courses into pathways. # 25 February 2025 ## Bug Fixes * Fixed course load failures due to thumbnail regeneration. # Infrastructure Changes: 18 February 2025 ## Enterprise Server Deployed Dedicated delayed-release server for Enterprise Headless. # 16 February 2025 ## Bug Fixes * Fixed Document-to-Course generating only 5-6 screens. # 11 February 2025 ## Bug Fixes * Sub-item navigation and mobile legibility fixes. # 08 February 2025 ## Bug Fixes * Fixed Arabic course generation issues. # 06 February 2025 ## Bug Fixes * OpenGraph tag fix. * Thumbnail not updating fix. * Quiz question visibility fix. * SCORM object load fix. # Coassemble 1.6.1 **21 January 2025** ## Onboarding Improvements Streamlined onboarding for new users. # Coassemble 1.6.0 **08 January 2025** ## Bug Fixes * Bulk invite issues fixed. ## Improvements * Removed English labels from Quizzes and Flashcards. * Test mode for payment links. * Onboarding improvements.