Tools
Check brand health from any connected agent.
Agents connected over MCP, including 
Claude, now read your brand health score and the text of every field you've filled in. Missing fields are listed by name, so agents know what to ask for.
Product updates
Brail product updates are done agentically every morning at 05:00: A Claude routine fetches the Github changes + calls 
Brail for Messaging, Brand voice and PlaybookProduct Updates, then writes the release note and publishes here.
UX/UI
Step through assets in the preview.
Previous and Next stay visible in the asset preview, with the end of the list dimmed.
UX/UI
See which files you just uploaded.
The upload-complete dialog now lists the name of every file you uploaded, so you can check the batch at a glance.
UX/UI
See how many assets match.
The asset library shows an exact count under the title, or a capped count when a search hits its limit.
UX/UI
Search the asset library as you type.
Results update as you type, words inside filenames now match, and your search stays in the page address.
Improvement
Filter the asset library by family.
Narrow the asset library to a single family to find related assets faster.
UX/UI
Confirm asset deletes before they happen.
Deleting an asset now asks first and tells you whether the hosted file is removed too, so nothing disappears by accident.
UX/UI
Act on selected assets faster.
Selected assets get their own action bar, and an empty library tells you what to do next.
Improvement
Faster dashboard pages.
Feel, Strategy, Voice, Messaging and the asset library now load and respond faster, and approving several assets at once happens in one go.
Tools
Upload files from disk to Brail over MCP.
Agents connected over MCP can now upload a file straight from disk to Brail, with no public link or base64 needed. Uploads stay pending until a human approves them, so you stay in control.
Tools
Start every brand task from one brief.
Agents now begin with get_brief, which picks the right playbook, template and brand context and returns the steps to follow, with your pinned inspiration images listed first.
Improvement
Sharper asset previews.
Asset previews are now generated at a higher resolution, so images look clearer when you review them.
Tools
Let agents fill in missing asset details.
Agents can now add captions, descriptions, alt text and tags to assets that have none, including pending uploads. Details you've already written are never overwritten, so your edits stay.
Feature
Link inspiration images to playbooks.
Publish a PNG, JPEG, WebP or GIF as inspiration and attach it to the playbooks it suits. Each playbook uses only the images linked to it, and the image stays general inspiration for the brand.
Improvement
Get principle suggestions from fewer images.
On the Feel page, one matching image is now enough to suggest a principle when you have seven or fewer unlinked images, and two when you have more. It used to take three. Suggestions stay drafts you review before saving.
Improvement
Switch feel principles on and off.
Turn individual brand feel principles on or off on the Feel page. Switched-off principles stay saved, so you can bring them back any time and stay in control.
Feature
Show brand feel with inspiration images.
Upload inspiration images on the Feel page to show what a habit looks like. Principles cite them, and Brail suggests a new one when three or more images share a habit. Agents can fetch them. Removing an image removes its principle links.
UX/UI
Activity log shows display names
The activity log now shows display names instead of agent names, so each entry is easier to attribute.
Improvement
Update the family of many assets at once
Select several assets and change their family in one operation, instead of editing each asset separately.
Improvement
Headlines and text are measured against the slot
Agents and the canvas now measure real font widths. A line that doesn't fit is rejected on save, a size under the minimum warns but still saves, and validate_creative checks without saving.
Feature
Brand feel principles for agents
Save an ordered list of feel principles on the new Feel page, with why each exists and what it looked like. Agents use them where your rules are silent; playbooks, shells and MUST rules still win.
Improvement
A chosen logo corner now places the logo
Choosing a logo corner now moves the logo, even when a gap or move-together rule includes it. A corner outside the safe area is reported instead of being ignored.
Feature
Warning when a logo sits on the wrong surface
The creative editor now warns when a light or dark logo variant sits on a clearly opposite background. Mixed photos and logos approved for either surface stay quiet, so the warning only appears when it matters.
Improvement
Text saves when you leave the field
Headline, text, brief and post text save as you move on from a field, and closing the page saves anything left over. A photo that meets the artboard edge is treated as a bleed and passes the safe-area check.
Improvement
Spacing follows the line actually shown
Single-line headlines no longer reserve space for a second line, so the field below sits the required distance under the text shown on the page.
Tools
Preview and edit draft creatives in chat
Draft creatives open as an inline preview in chat. Edit the headline, subheadline and caption in place, swap in an approved photo and save. Open in Studio for the full editor.
Feature
Let a shell choose its starting logo
A shell's logo field can now set the logo a new creative starts with and limit which logos that shell allows, so creatives begin with the right mark.
Tools
Browse approved images in a chat gallery
Ask to see boat photos or logo versions and a gallery of your approved assets opens in chat. Each thumbnail shows whole, with a Download (original) button for the stored file.
Improvement
Brand fonts load in the studio editors
Fonts from your brand's catalogue now load in the Studio editors, so text on the canvas appears in the typeface your brand uses rather than a fallback.
Tools
Agents can draft in Studio and pick photos
Agents can list your shells and drafts, place or revise text on your own drafts and choose an approved photo. Your approved creatives and shell structure stay untouched.
Feature
Start Studio from a chosen shell
Studio's home now shows your brand's shells. Start a creative from a shell, a brief or both, and an empty brief opens a blank draft on that shell.
Improvement
Separate caption step for social creatives
Shells with a companion caption now edit the image and post text separately, with the caption written from your brief and the lines on the artwork.
Feature
Crop a photo inside its frame
Creatives can zoom and pan a photo inside its frame while the frame stays put. Shells choose whether crop is on and how far it zooms, so you stay in control.
UX/UI
Undo, redo and nudge in the shell editor
Undo and redo now cover shell edits, with a drag counting as one step. Arrow buttons move the selected field by one pixel, including a box that hugs the fields inside it.
Feature
Start a blank shell from the format catalogue
A new shell can now open as a draft for any catalogue format, sized to that format's artboard, alongside copying a shell or starting from a file.
Improvement
Shell type and gaps scale with the format
Type sizes and keep-a-gap distances now scale with the artboard, so a shell keeps its proportions when you change format, and returning restores the original sizes.
UX/UI
Round the corners of a shell box
Shell boxes can now have rounded corners, set in pixels in the editor. The radius carries over when you switch format, and it shows in both the preview and the creative canvas.
Feature
Place a layout into a shell
Reuse a saved layout in another shell by sizing a rectangle and placing it. Later edits to the layout leave shells that already used it unchanged.
UX/UI
Edit the sample text on a shell
The headline and subheadline in the shell editor can now be typed in the toolbar and are saved with the shell. Shells created without that copy still open on the short sample.
Improvement
Each brand gets its own shells
Shells now belong to a single brand, and a creative only uses the active shells of its own brand. Existing shells have been moved to the brand they belong to.
Improvement
Choose brand colours for a text field
In the shell editor, you can now select which brand colours are available for a text field, so creatives only pick from the colours you have approved.
Feature
Adapt a shell onto another format
A new Another format action beside Save copies the current layout, including unsaved edits, onto another format as a draft. The original shell is left unchanged.
Feature
Upload HTML templates to playbooks
Upload an HTML template on a playbook page and connected agents can list it and download it on request. Context packs stay lightweight.
Improvement
Text shrinks to fit its box
Text now fits its box automatically: type steps down until the words fit, and you only see a note when it adjusts. Opt-in per slot; you stay in control.
Improvement
Feed post kept off the image
On-image lines are now short headlines with no caption or hashtags. Media formats get a separate Caption field under the artboard for the feed post. Text posts and stories are unchanged.
Feature
Playbook revision history
Playbooks now keep a revision history. View earlier versions on the playbook page and restore any of them, so nothing is lost and you stay in control of what's active.
Tools
Generate on-brand copy and brand context over HTTP
REST endpoints now generate on-brand copy and fetch brand context packs over HTTP, previously MCP-only. Authenticate as a brand member; no service-role key needed. Calls log to your activity feed; prompts and copy are never stored.
Improvement
Text-only drafts no longer fail on layout rules.
Wordmark, headline accent, and illustration rules now run only through visual compliance checks. Text-only drafts clear on copy rules alone, so correct copy ships without being blocked.
Feature
Exemplars scoped to audience personas.
Save a message pattern once and reuse it for a persona across every channel, instead of duplicating it per playbook. Available via MCP, context packs, and the Exemplars dashboard.
Tools
Agents fetch asset files directly.
Claude, Cursor, and other connected agents now retrieve an asset's original file directly through get_asset, up to 8MB, instead of only a signed URL.
Improvement
Draft with Brail keeps your work in view
Your draft stays open until you save each section or discard it, so one section saves without losing the rest. Slow drafts show a clear timeout message instead of a generic error.
UX/UI
A cleaner playbook header
The draft/live toggle now sits in the playbook header, with the back link, description, status badge and draft banner removed. Toggle and button click bugs are fixed.
Feature
Reopen a rejected proposal
Rejected proposals can now be reopened for review, so there's no need to resubmit from scratch.
UX/UI
One consistent header style across Brail
Settings, assets, exemplars, playbooks, activity, proposals and changelog now share one header style, with content aligned to the sidebar.
UX/UI
Clearer section labels
Strategy, messaging, voice, governance and visuals pages now use pill-styled section labels with icons, making each page easier to scan.
Improvement
Brand suggestions render as readable text
Brand suggestions and audience-message mappings now render as readable text in proposal and live cards, instead of raw JSON or [object Object].
Tools
Agents preview assets without extra steps
Connected agents now get a ready-to-view thumbnail image when they fetch an asset over MCP, instead of needing extra flags or the full original file.
UX/UI
See everything on a card without opening edit
Cards with long lists now expand to show everything in place when you click "+N more", instead of sending you into the edit view.
Improvement
Fixed duplicate logo entries
Removing a logo variant no longer leaves duplicate entries behind in your asset library.
UX/UI
Voice vocabulary proposals display correctly
Suggested vocabulary and jargon substitutions in AI proposals now display as readable text instead of raw data when you review them.
Improvement
Visual compliance checks catch more real logos and type
Compliance checks now read your actual typography and recognise wordmarks picked up by OCR, cutting false missing-logo and font-mismatch flags on real designs.
Improvement
Replace a font file without starting over
Upload a new file for an existing font instead of deleting and re-adding it. Metrics are now pulled automatically for fonts linked from Google Fonts and Fontshare, not just uploads.
Improvement
Check compliance against a specific playbook
Choose which playbook to check content against, instead of your brand's default rules, so results match the channel you're actually shipping to.
Connector
Accurate font metrics via MCP, no font file needed.
AI agents connected via MCP now receive precise OpenType metrics for your brand fonts, so text layout works correctly without sharing the binary font file.
Feature
Upload your own fonts.
Add WOFF2 or OTF files directly to your brand's typography settings. OTF files convert to WOFF2 automatically. The font previews instantly so you can confirm it renders correctly.
Improvement
Smarter context for AI agents.
Agents connected to Brail now retrieve brand exemplars for visual design tasks only. Copywriting tasks receive focused verbal context instead.
Feature
Give each playbook its own exemplar library.
Exemplars now attach to individual playbooks instead of one shared channel, so each playbook, including a dedicated one like LinkedIn, keeps its own library.
Improvement
PDF assets now show a thumbnail of the first page.
Upload a PDF and see a thumbnail of its first page in your asset library, so you can recognise files at a glance instead of guessing from the filename.
Improvement
Assets now land in the right family automatically.
When enrichment classifies a new asset as a logo, photo or icon, it sets the family for you — without overriding one you've already chosen.
Improvement
Skip or cancel website setup when creating a brand.
Finish brand setup without crawling a site, or cancel the flow entirely — no longer forced to retry or complete a website step.
Feature
Save exemplars at the brand level.
Exemplars no longer need a channel — save visual recipes, on-brand references, and other examples directly at the brand level.
Improvement
Colour usage rules now parse correctly.
Paste colour rules with names or hex codes across multiple lines — Brail matches each to the right palette colour automatically.
Improvement
Channel playbook skills field raised to 40,000 characters.
Write as much channel-specific guidance as your agents need — the skills field now holds up to 40,000 characters, up from 8,000.
Feature
Define colour steps and bind them to semantic colour roles.
Build a tint or shade scale from any base colour in Visuals, then bind each step to a semantic colour role alongside your Identity colours.
UX/UI
Changelog audit trail gets a cleaner layout.
Changes by pillar now show as large stat numbers instead of a plain list, with a clearer header up top, so the audit trail is easier to scan.
UX/UI
See your teammate and their agent side by side in the activity log.
Each activity log entry now pairs your teammate's avatar with the connected agent's logo, and shows the full date and time, not just the time.
UX/UI
Typography styles link directly to your brand fonts.
Each typography style now binds to a specific brand font, with an optional letter-spacing override that falls back to the font's default. See exactly which font powers every style.
Improvement
Track thumbnail and enrichment jobs from any asset.
You can now see exactly where any thumbnail or enrichment job stands — queue position, elapsed time, and a clear error if something fails — right from the asset detail view.
Improvement
Plans simplified to Free and Paid.
Two plans now: Free and Paid. Need more? Request a quote straight from your subscription settings.
Connector
Design tokens for connected agents now include fonts and logo rules.
Agents connected via MCP — including Claude and Cursor — now pull fonts, colour usage rules, layout principles and logo rules from your design tokens automatically. No extra briefing needed.
UX/UI
Strategy, Visuals, Messaging and Voice pages get a cleaner header
The Strategy, Visuals, Messaging and Voice pages now open with a clearer, more compact header, so you can see the page context and jump into editing faster.
Feature
Make asset thumbnails publicly viewable
You can now choose to make an asset's thumbnail publicly viewable by URL, useful when a connected tool needs a direct link. Off by default; you stay in control from brand settings.
Improvement
Cut-off image uploads are now caught before they save
Uploading a truncated or corrupted image file, via the dashboard or MCP, now fails clearly with guidance to re-export and retry, instead of silently saving a broken thumbnail.
UX/UI
Connectors page redesigned for easier scanning
The connectors page is simpler to scan, with clearer connection status, a redesigned setup guide button, and refreshed card copy across Claude Desktop, 
ChatGPT, 
Figma and more.
Improvement
Connector setup guides refreshed for accuracy
Setup guides for Claude, ChatGPT, Figma, 
Lovable, v0, Cursor, Slack, Notion, 
Microsoft Copilot Studio and more now match each tool's current connection flow, so you can connect Brail correctly on the first try.
Connector
Claude Design added as a connector
Claude Design is now available as a connector, alongside Claude Desktop and ChatGPT, so you can pull approved logos, imagery and copy exemplars from your library straight into your designs.
UX/UI
Asset ID now shown in the metadata tab
Every asset's ID is now visible in its metadata tab, so you can reference the exact asset when working with support or the API.
Connector
Updated logos across your connections page.
Claude, ChatGPT, Microsoft Copilot, Perplexity, Cursor, Figma, Lovable, Notion, v0 and Windsurf all show refreshed logos, so you can spot the right connector at a glance.
Improvement
Suggest improvements now applies to prose fields.
Suggest improvements now applies to prose fields only. Review each suggested edit before it applies — you stay in control.
Tools
Regenerate an asset's thumbnail on demand.
Open any asset's detail view and regenerate its thumbnail whenever it's incorrect or outdated, including fixes for Figma icon thumbnails that weren't rendering correctly.
UX/UI
Read full field text without opening edit.
Read full text inline in strategy, messaging, voice and governance fields, with an expand toggle for longer entries — no need to open edit first.
Improvement
Fact grounding rule edits now show a warning first.
Changes to fact grounding rules affect what agents and compliance checks treat as verified. Brail flags this before you save, so you review before it takes effect.
UX/UI
Edit governance rules in the same layout as strategy, messaging and voice.
Fact grounding rules, geographic restrictions and product name rules now use the same editor layout as strategy, messaging and voice fields.
Improvement
Control when sources refresh.
Brail no longer recrawls sources automatically — you decide when to pull the latest version. Trigger a re-crawl on any source whenever you want it refreshed.
Connector
Connect Figma through MCP.
Import from Figma and push design-system updates via MCP, replacing the previous plugin and OAuth connection — one consistent way to bring Figma into your brand.
Improvement
Track plan usage from subscription settings.
AI credits, seats, and assets are now visible in one place under subscription settings. Soft warnings fire before you hit a limit; hard caps pause new usage. Nothing already saved is removed or locked.
UX/UI
Filter your activity log by teammate, agent, or date.
Events are now grouped by day. Filter by teammate, AI agent, or date range, or search the full log — each event shows the asset, the change, and who's responsible.
Improvement
Visual compliance checks now read prose-style colour usage rules.
Visual compliance checks used to fail on colour usage rules written as free text. They now read prose-style rules correctly, so checks run reliably either way.
Improvement
Font usage notes now reach every agent automatically.
Font usage notes — letter-spacing, placement, weight — now reach every connected agent automatically. No extra lookup needed to stay on-brand.
Connector
Connect Notion to your brand
Notion now reads your live brand context via MCP, so pages you create there start on-brand automatically.
UX/UI
Broken links now show a branded 404 page
Signed-in visitors get a way back to their brand; signed-out visitors get a way to get in touch, instead of a generic dead end.
Improvement
Wordmarks and symbols now show up when adding logos from your library
Wordmarks and symbols now appear alongside primary logos when you add one from your library, so you pick the right variant right away.
Connector
Import your brand directly from Figma
Connect a Figma file under Sources to pull colours, typography and spacing straight into your brand, ready to use immediately.
Connector
Connect v0 to your brand
v0 now reads your live brand context via MCP, so everything you build there is on-brand from the first prompt.
Connector
Connect Slack to your brand
Slack now reads your live brand context via MCP, so messages and content drafted there come out on-brand automatically.
Tools
Bigger files now upload cleanly via MCP
Agents can point to a public link for an asset instead of embedding it, so large files reach Brail without passing through the model context.
Connector
Windsurf connects to Brail via MCP
Windsurf now reads brand-accurate context directly from Brail via MCP, keeping outputs on-brand automatically.
UX/UI
Redesigned connections page
Set up any connector faster with a clearer layout, step-by-step guidance for each tool, and a self-serve form to request one that isn't listed yet.
Improvement
Infinite scroll in the asset library
The asset library now loads more assets automatically as you scroll, replacing manual pagination, in both the asset registry and the public asset hub.
UX/UI
Redesigned proposal review.
Reviewing proposed brand changes now has a clearer layout, easier-to-read field-by-field diffs, and simpler accept and reject actions.
UX/UI
One consistent confirmation toast across Brail.
Actions across proposals, sources, assets, exemplars and the console now confirm the same way, so you always know when something has saved.
Connector
Connect Lovable to your brand.
Lovable now pulls your live brand context, so what you build there starts on-brand.
Improvement
Fixed asset search results on mobile.
The asset library now displays search results correctly on mobile — tiles, thumbnails and detail views are properly sized and spaced.
Improvement
Control how often sources recrawl.
Choose whether and how often each source recrawls for changes. Off by default; you stay in control of when content updates.
Connector
Every connected agent now gets full MCP tool details.
Claude, ChatGPT and other connected agents now receive the full tool details for three Brail MCP tools, so calls to submit brand suggestions, upload exemplars and check visual compliance work as expected.
Improvement
Visual compliance no longer auto-attaches exemplars.
Automated visual compliance checks no longer pull in exemplars by default. They remain available as creative reference for your agents.
UX/UI
Redesigned exemplars registry and detail view.
The exemplars registry and detail view now use a cleaner, more structured layout, so you can review and edit examples faster.
Improvement
Exemplars no longer expire automatically.
Exemplars no longer need a start or end date. Add one and it stays active until you archive it yourself.
Feature
Download an exemplar's original file.
You can now download an exemplar's original file straight from its detail view, no need to screenshot or request it separately.
Improvement
Exemplar uploads from agents now land correctly.
Exemplars uploaded by agents — by file URL or directly — now appear correctly in your Exemplars dashboard.
UX/UI
Brail's icon in AI tools now shows clearly on dark themes.
The Brail icon in Claude, Cursor and other MCP clients previously appeared faint on dark themes. It's now a solid, high-contrast tile everywhere.
Feature
Agents now run visual compliance checks through MCP.
Claude, Cursor and other connected agents now run visual compliance checks on their creative work, with your logo usage rules and playbook guidance enforced automatically.
UX/UI
Transparent access when connecting agents.
Connecting Claude or Cursor to Brail now shows exactly what access is granted, matched to your Brail role. Revoke access anytime in Brail.
Improvement
Sign-in verification only when needed.
Brail now asks for extra verification only after suspicious sign-in activity. Normal sign-ins are unaffected, and you stay in control of your account security.
UX/UI
Sign-in screens redesigned.
Login, sign-up, forgot-password, and accept-invite screens now share a consistent, simplified layout.
Feature
Preview assets full screen.
Open any asset full screen and move to the next or previous one without leaving the view.
Connector
MCP tool results now display in Cursor.
Brail's MCP tools now include readable text alongside structured data, so brand lookups display correctly in Cursor and other MCP clients.
UX/UI
Clearer explanations when an agent can't connect.
The screen shown when Claude, Cursor, or another agent can't connect to Brail now explains exactly why, and what to do next.
UX/UI
Sidebar mark now opens your brand home directly.
Clicking the Brail mark in the sidebar now takes you to your selected brand's home, instead of the brand picker.
Improvement
Enrich with AI hides once an asset is already enriched.
The action no longer appears in the detail view after an asset has been enriched, keeping available actions relevant.
Improvement
Copy URL removed from the asset detail view.
Signed URLs expire within an hour, so the copy-link action has been removed to avoid sharing links that stop working.
Improvement
Assets download straight to your device.
Downloading an asset from the library now saves the file directly to your device, instead of opening it in a browser tab.
UX/UI
Update visual rules faster.
Colours, logos, fonts, typography, spacing and layout principles now live in dedicated, sectioned editors within brand visuals.
Improvement
Invite and request-help emails now match your brand.
Invite and request-help emails now arrive in one branded template, with matching HTML and plain-text versions every time.
UX/UI
General settings: owner-only access.
General settings now opens for brand owners only; other members are redirected to Members with no change to their other permissions. Agent permissions, asset enrichment and hub visibility sit in a clearer layout.
Improvement
Faster invite acceptance for existing users.
Existing users accepting a team invite now do so in a single step, instead of moving through separate screens.
Feature
Check text and design compliance in one workspace.
Upload .txt, .md or .docx files, review flagged issues inline, and export a clean record to Markdown or print.
UX/UI
API keys, members and billing settings redesigned.
Manage API keys, invites and your subscription without switching between separate pages — everything now shares one consistent layout.
UX/UI
Redesigned asset detail view.
Asset details now open in a structured layout — overview, metadata, usage and actions — so you find what you need and act on it faster.
UX/UI
Navigate brand settings faster.
Brand settings now use the same tab layout as the rest of Brail: General, Members, Subscription, API keys and Integrations, all in one row.
Feature
Marquee select in the asset library.
Drag a selection box across tiles to multi-select assets for bulk actions. Hold Shift or Cmd to add to an existing selection. Checkboxes still work as before.
Improvement
Fixed extra scroll space in search results.
Asset search results no longer have extra scroll space below them — a hidden layout element causing the issue has been removed.
Feature
Download multiple assets at once.
Select assets in your library and download them together as a zip — no more downloading one at a time.
Improvement
Colour swatches in Ask replies.
Ask now shows a colour swatch next to every hex code in its replies — and when the colour matches a brand token, it displays the token name too.
UX/UI
Cleaner asset tiles, with details on hover.
Tiles now show a clean thumbnail by default. Title and format badge appear on hover or keyboard focus. On touch devices, both stay visible always.
Feature
Choose what happens right after you upload.
After uploading assets, pick your next step — edit metadata now, approve everything as-is, or leave it for later, whichever suits your workflow.
Improvement
Asset details no longer autosave.
Editing an asset's details no longer autosaves on every change. Hit Save when ready — a confirmation prompt appears if you navigate away with unsaved changes.
Improvement
Activate a draft channel straight from its banner.
Draft channels now show an Activate channel button in the banner. Teammates without edit access see a note that only editors can change channel status — access rules stay as set by your team.
Improvement
Upload larger PDFs without hitting the limit.
PDF sources now upload directly to storage, so file size is no longer a barrier when you add a source.
Improvement
Upload assets and watch progress in real time.
See live, per-file progress when you upload to your asset library, with support for concurrent uploads. Assets upload directly, with no extra step to add a reference first.
Feature
Source detail view shows status and outcome at a glance.
Open any source to see its detail view, current status, and a flag when it produced no proposals, so you can see exactly what each source produced.
Feature
Reset your password without waiting on support.
Reset your password from the sign-in page. Request a reset link by email, verified with hCaptcha, and set your new password straight away.
Improvement
Quick-setup images now live in your asset library.
Logos and images imported during quick brand setup are saved directly in your asset library, kept in place regardless of changes to the source page.
Improvement
New teammates go straight to account creation.
When you invite someone without a Brail account, they now see Create account only, with no confusing Sign in option. Existing users still sign in as before.
UX/UI
Library now opens straight to your assets.
Clicking Library in the sidebar now takes you straight to your assets, skipping the extra landing step.
Improvement
Continue onboarding even with no social profiles.
Removing all social handles during quick setup no longer blocks you from continuing; onboarding continues without them.
Improvement
Compliance checks are softer on soft claims by default.
Compliance checks no longer flag soft claims as errors by default. Turn on strict mode for any channel playbook, such as thought leadership articles, for tighter fact-grounding.
UX/UI
"Click to add content" now opens the editor, like Edit.
Verbal and Strategy: clicking "Click to add content" now opens the same editor as Edit, so both paths behave the same.
Feature
Brand Assistant now answers questions on your public hub.
Your public brand hub now offers Brand Assistant, so visitors get answers about your brand directly. Off by default, you turn it on in General settings.
Improvement
Ask sessions stay private, nothing saved locally.
Ask sessions are stateless. Chat transcripts no longer save to your browser's local storage, so nothing persists between sessions.
UX/UI
Ask opens with starter prompts.
Ask now opens with a set of starter prompts, so you can jump straight into a question about your brand.
Improvement
Request help is now a single modal.
Requesting help from a teammate now happens in one modal — pick the pillars you need, choose a teammate, and Brail sends the email. No link to copy, no address to type.
Feature
Publish your brand as a shareable, read-only site.
Share your logo, colours, typography, strategy and assets at your own link. Set visibility to private, link-only, or public — private by default, so you stay in control.
UX/UI
Fill brand gaps in a conversation, not a form.
Fill the gaps now runs as a chat — same interface as Brand Assistant, with message bubbles, a typing indicator, and markdown replies. Leaving with unsaved answers now triggers a warning.
Improvement
Voice suggestions for vocabulary, writing rules, and technical rules now save reliably.
Suggested changes to vocabulary, writing rules and technical rules — including their notes — used to fail to save, sometimes silently. They now save reliably, with clearer errors on failure. You review and approve every change.
Improvement
Typography preview now loads your brand's actual fonts.
The live typography preview now loads your brand's actual font files instead of just applying the font name, so it shows what your typography really looks like.
UX/UI
The vocabulary list in brand Voice settings now stacks vertically.
The preferred, banned, and domain-jargon word lists used to display as three columns side by side, which made them hard to read. Each list now stacks vertically, making the words easier to scan.
Feature
Layout principles, typography, and value propositions now start from a generated first draft.
Generate draft now creates a first version of layout principles, a typography scale, and value propositions from your existing brand data, so you edit a draft instead of starting from a blank field.
Improvement
Compliance checks now succeed reliably on longer drafts and lightly formatted AI responses.
Compliance checks used to fail on long drafts or lightly formatted responses. They now scale to the draft's length, read lightly formatted responses correctly, and retry once if needed.
Feature
Compliance checks now flag generic, AI-sounding writing.
Compliance checks now also flag generic AI writing — marketing puffery, overused AI vocabulary, vague claims, and repetitive patterns like em dashes — alongside your brand voice and governance rules.
Improvement
Segment → message mapping displays correctly everywhere.
Segment-to-message mapping content could be stored as a raw object when it came from the AI co-editor or an approved AI proposal, so it showed up garbled in the brand hub, exports, and prompts sent to connected agents. The mapping now always renders as readable text, and any brands affected by the earlier storage format have been corrected.
Improvement
External agent prompts now respect your marketing channel setup permission.
The copyable prompt for Claude, Cursor, ChatGPT or another LLM connected to your Brail MCP server used to describe a fixed set of permissions on the channel page, regardless of what you'd chosen under Settings → General → Marketing channel setup. It now reflects your actual setting: under Always, the prompt tells the agent it can write directly to both draft and active playbooks; under Draft only, it tells the agent that active channels are read-only for MCP writes and directs it to suggest changes for you to review and save in Brail; with marketing channel setup switched off, the option to draft with your preferred LLM no longer appears on the channel page. You stay in control of what gets written to your live channels.
Improvement
AI agents now fetch your real logo instead of inventing one.
Claude, Cursor and other agents connected over MCP had a binding fact-grounding contract for copy, but not for logos and brand marks — an agent building a visual deliverable could approximate or invent your logo when the exact file wasn't at hand. get_brand_context_pack now returns a visual grounding contract for visual, assets and design system tasks, together with asset hints that name the exact tool an agent must call — get_asset or get_assets_batch — before using any mark, and requires it to say so explicitly when no matching asset exists rather than fabricate one. This applies automatically, with no setup required on your part. A new resolve_assets option also returns your default logo in the same call, for agents that just need the primary mark.
Improvement
Agents connected over MCP can now act on your brand as soon as they connect.
Claude, Cursor and any other MCP-connected agent can submit brand suggestions, upload assets and exemplars, and set up or update marketing channel playbooks without you changing a setting first. All four permissions now default to on for every brand, including brands that existed before this change — and you stay in control, with each one you can switch off individually in Settings → General.
UX/UI
A calmer top bar and consistent navigation everywhere.
Every section of Brail — Brand, Playbooks, Activity, Library, Admin, Agents and Settings — shares the same pill-style sub-navigation with icons now, so switching areas feels the same everywhere. Search sits centred in the top bar on desktop and drops to its own row below the navigation on tablet and mobile, so it never crowds against the pills or your account menu.
Improvement
Compliance checks now accept your verified connector facts.
check_compliance and compliance_final_copy previously flagged accurate connector details — a tool's URL, its OAuth support, its features — as unverified, because the standalone check only saw your brand model, not the brief behind the copy. Both tools now accept an optional grounding_brief parameter, so verified facts pass through and compliance checks return an accurate, publishable result.
Improvement
Choose how MCP agents update your marketing channel playbooks.
Settings → General → Marketing channel setup now offers three levels of access instead of a single on/off toggle: Always lets agents create channels and update both draft and live playbooks; Draft only limits agents to draft playbooks, with live channels only getting read-only suggestions to review and activate; Off blocks all MCP playbook writes. Pick the level that fits your team — you stay in control.
Feature
MCP agents can now upload post exemplars.
Let your Claude or Cursor agents log what good and bad posts look like — uploads land as drafts until a human publishes them in the Exemplars dashboard. Off by default; you stay in control.
UX/UI
Switch channel status with one toggle.
Activating or moving a channel to draft is now a single switch on the channel and playbook page, so you can see and change its status at a glance.
Improvement
Segment-to-message mapping boxes now save correctly on reload.
Per-segment messages in the segment-to-message mapping editor could collapse into a single free-text box after saving and reloading, especially when a message spanned multiple lines or sentences. Each segment's message now round-trips correctly, so your mapping stays exactly as you wrote it.
Connector
Brail's own icon now appears in connected AI tools.
The Brail MCP connector displays Brail's own icon instead of a generic default glyph, so you can spot the Brail connection at a glance across any MCP-compatible AI tool you use.
Feature
Governance area surfaces your fact-grounding rules.
A new Governance area, reachable from the brand menu, puts your fact-grounding rules front and centre. Compliance checks now flag claims in AI-drafted copy that aren't backed by your brand's own documented facts, and flags surface as warnings — you keep full control over what publishes.
Feature
AI skill suggestions for playbooks.
Get a ready-to-use skill suggestion for any playbook's Skills field, grounded in your brand data and that playbook's own strategy, audience and KPIs. Review the suggestion and its rationale, adjust as needed, then save.
Feature
Skills field on playbooks.
Write free-form instructions directly on any playbook's Skills field. Brail delivers them straight to every connected agent drafting on that channel, ranking below compliance and MUST rules so your guardrails always hold.
Improvement
Sign in without interruption on hCaptcha-protected accounts.
Password sign-in is unblocked for accounts with hCaptcha protection enabled — a captcha error that blocked login is resolved. The sign-in page also renders the captcha at its natural size, and the Google and Microsoft buttons now match Brail's own button style.
Feature
Create playbooks for any channel, not just our built-in list.
Marketing is now called Playbooks — existing playbooks are unchanged, only the name has moved. A new Other type lets you add a playbook for any channel outside the built-in catalogue, such as case studies or newsletters, each with its own dedicated page.
UX/UI
Brand showcase page decluttered.
Drop shadows and the brand-scoring panel are gone — the showcase page now has a flatter, cleaner layout, making your brand presentation easier to scan and share.
UX/UI
See every channel's status at a glance.
The marketing channels list now shows each channel's status, so you can tell what's set up and what still needs attention without opening a single one.
Feature
Let agents set up your channels.
Connected agents can now create and configure your marketing channels — filling an empty field directly, or suggesting a change when one's already set, so nothing you've written is overwritten without your say-so.
Feature
Let agents add to your asset library.
A new brand setting lets your agents upload assets straight into Brail — so the work they produce lands where your team can find and reuse it. Off by default; you stay in control.
Improvement
Press releases get their own channel.
Press release is now a channel in its own right, with its own format and playbook — so announcements read like announcements, not repurposed social copy.
Improvement
Brand-suggestion controls, where you'd expect them.
The setting that governs how agents suggest changes to your brand now lives in Settings → General, alongside your other brand controls.
Improvement
Agents pinpoint the exact brand data.
Agents can now find the precise content pillar and field they need in your brand model — so the context they pull back is exact, not approximate.
Improvement
One language for audiences, everywhere.
Your audience model now uses one consistent name — segments — across the app, the API and every connected tool. No mismatched labels, no confusion between Brail and your stack.
Feature
Start a brand over with one click.
Need a clean slate? You can now reset a brand and rebuild it from scratch — without deleting the workspace or losing your team.
Improvement
Agents act on your proposal feedback.
When you approve or flag a change an agent suggested, that verdict now flows back to the agent — so it learns what your brand will and won't accept.
Improvement
The assistant suggests only when you want it to.
The Brand Assistant now offers to make changes only when you signal you're contributing — so a quick question stays a question, not an interruption.
Improvement
Stronger protection on sign-up.
New sign-ups are now shielded by bot protection, keeping fake accounts out without slowing your real team down.
Feature
Sign in with Microsoft.
You can now sign up and log in with your Microsoft account, alongside Google — fewer passwords, faster access for your whole team.
Improvement
Sharper instructions for your agents.
We refined the guidance Brail gives connected agents, so the copy and context they pull back is more accurate out of the box.
UX/UI
Formatted replies in the assistant.
The Brand Assistant now renders headings, lists and emphasis properly, so its answers read like a document, not raw text.
Improvement
Filter your examples.
Find the right example faster — the exemplar library now filters down to exactly what you're looking for.
UX/UI
A calmer loading experience.
Pages across Brail now load with a quiet, on-brand animation instead of a blank wait — so moving around feels instant.
Improvement
The assistant edits only what it should.
The Brand Assistant now proposes changes only to your strategy, voice and messaging — the places that should be edited in conversation — and leaves everything else untouched.
Improvement
Security hardening across the platform.
A round of security patches and dependency updates — quietly keeping your brand data and account safe.
Connector
Connect your brand to Claude Design.
Designers in Claude Design now pull live brand context — colours, typography, logos and approved assets — straight from Brail, with guardrails applied as they generate. No manual setup, no stale exports.
Improvement
Better brand colours from your website.
Onboarding now reads your site's colours more accurately, so your palette lands closer to right from the very first scan.
Feature
Usage rules for every asset.
Set how each logo, image or file may be used, and let Brail enrich assets with on-brand metadata — so the right asset is always easy to find and never misused.
UX/UI
Sources, redesigned.
Everything that trains your brand now lives in one clean list. Link a URL or upload a file, choose how often each source re-crawls, re-sync on demand, and see exactly which parts of your brain each source feeds. No more separate tables, no clutter.
UX/UI
Sharper page titles and favicon.
Every page now carries its own title and the Brail mark — so saved tabs, bookmarks and shared links are instantly recognisable.
Improvement
LinkedIn Articles get their own channel.
LinkedIn Articles are now a channel in their own right — with their own format and playbook — instead of being squeezed into one generic LinkedIn setup.
Feature
Delete a brand workspace.
You can now remove a brand workspace you no longer need, cleanly and completely.
UX/UI
Brand Assistant, front and centre.
We renamed Chat to Brand Assistant and stripped the redundant headers off your hub pages — so every page gets you to the work faster.
Feature
Agents can flag gaps back to Brail.
When a connected agent hits something your brand doesn't cover yet, it can now report that gap — so you know exactly what to fill in next.
UX/UI
A lighter, calmer interface.
A refreshed neutral palette and a simpler loading screen across the app — easier on the eyes, with less visual noise between you and your brand.
UX/UI
A cleaner Training page.
Each training card now has its own clear action, and on/off choices read as simple green/red tags — so it's obvious what to do next to strengthen your brand.
Improvement
Tune which content pillars matter most.
Weight your content pillars by importance, so the work your agents produce leans on what matters most to your brand.
Feature
Marketing channels, rebuilt.
Channels are now unified — one Instagram, one LinkedIn — with sensible defaults, a setup Brail can draft for you, and the option to remove any channel you don't run. Everything your agents need to post on-brand, per channel.
Improvement
Auto-detect your brand's language.
Onboarding now recognises the language your brand writes in, so suggestions and copy come back in the right voice from the start.
UX/UI
Navigation cards across every hub.
A consistent set of navigation cards now guides you from each hub to the next step — so finding your way around is one obvious click.
Feature
Choose your AI model.
Pick the Claude model that fits the job — Opus for depth, Sonnet for balance, Haiku for speed — and let suggestions run on a model tuned for quick, on-brand edits.
Feature
Add sources straight from Training.
You can now link a URL or upload a file directly from your Training page — feeding your brand without leaving the page you're on.
UX/UI
A clearer activity log.
Your activity log now reads in plain language — easy to scan for what changed, when, and who did it.
UX/UI
A bigger, smoother assistant.
The Brand Assistant window is larger and scrolls cleanly, with a clear "Brail is thinking" indicator while it works.
UX/UI
Tidier Voice rules.
Your voice's technical rules — formatting, do's and don'ts — now sit in a single combined card with room for a note on each choice, instead of scattered fields.
Improvement
Crawl JavaScript-heavy sites.
Brail now reads brand content from sites that render in the browser — so modern, app-like websites import just as cleanly as static ones.
Improvement
A clearer audience model: Segments and Audiences.
We renamed and reshaped your audience model into Segments and Audiences, with each audience linked to its segment — a structure that's easier to reason about and reuse.
Improvement
Tidied brand fields.
A cleanup pass across your strategy, visuals and governance fields — clearer labels, less duplication, easier editing.
Improvement
Capture spacing and layout principles.
Your brand model now captures spacing, sizing and shadow principles alongside colour and type — so agents build layouts that feel like you, not just copy that sounds like you.
Improvement
Your agents can see your examples.
Approved and rejected examples are now visible to connected agents — teaching them what on-brand really sounds like, in your own words.
UX/UI
Redesigned Voice.
Brand voice gets the same treatment: tone, rules and vocabulary in tidy cards, with simple pickers for the set choices.
UX/UI
Redesigned Messaging.
Your messaging — pitch, value props, taglines, proof and a structured message hierarchy — now lives in clean two-column cards that are quicker to read and edit.
Feature
Map audiences to messages.
Connect each audience to the messages that land for them — so every agent knows what to say to whom.
Improvement
Import examples and images from social.
Pull your real posts and images straight from social media into your example library — with the post link filled in for you.
Improvement
Faster brand pages.
We trimmed unnecessary data fetching across your brand pages — so your overview, brands list and assets open noticeably quicker.
Improvement
Contribute from the assistant.
You can now submit proposals and add examples right inside the Brand Assistant — capture an idea the moment it lands, without switching pages.
Feature
A redesigned brand chat.
A faster, clearer conversation with your brand brain — ask anything in plain language, and drop in example files mid-chat to teach it as you go.
UX/UI
Redesigned Strategy.
Your strategy now lives in per-field cards — positioning, promise, audiences, values and more — with suggestions opening in a focused pop-up. Long entries and structured fields read cleanly instead of as raw data.
Improvement
Readable brand URLs.
Brand links now use a readable name instead of a long ID — tidier to share and easier to recognise.
UX/UI
Redesigned navigation and workspace switcher.
Cleaner second-level tabs, a warmer page surface, and a redesigned brand switcher that always fits its name — so moving between brands and sections is effortless.
UX/UI
New editing surfaces.
A right-hand slide-in editor and read-at-a-glance excerpt cards make editing brand content quicker and less disruptive.
UX/UI
A refreshed design system.
A ground-up refresh of Brail's look — a single colour scale, a clear type scale, and consistent cards, buttons and tabs everywhere. The whole app feels calmer and more coherent.
Feature
A new Training page.
A post-onboarding home for strengthening your brand: see how complete each part is, what's connected, and exactly which gaps to fill next — built from your real brand data.
Improvement
A better asset library.
Improvements to how your assets are stored, browsed and tagged — your brand's media, easier to find and keep on-brand.
UX/UI
A fresh icon set, and clearer names.
A crisper, more consistent set of icons across the app. "Agents" is now Tools and "Connect" is now Connectors — names that say what they do.
UX/UI
Redesigned new-brand onboarding.
Setting up a brand is now a calmer, split-screen flow with a guided card deck — less form, more conversation.
Feature
An in-app product tour.
A built-in tour walks you through Brail the first time, so you know where everything is from day one.
UX/UI
A welcome moment after setup.
A short slideshow celebrates your finished brand setup and points you to what's next.
Feature
Your brand report after onboarding.
Finish setup and get a clear report of what Brail found and built — your brand, summarised back to you.
Improvement
Better font and style detection.
Onboarding now reads the fonts and styling from your website more reliably — so your brand looks right sooner.
Feature
Marketing channels and playbooks.
A new Marketing layer where each channel carries its own framework and playbook — the formats, rules and guidance agents follow to post on-brand.
UX/UI
Redesigned connections page.
Your connections now sit in clean, uniform cards — a clearer view of everything wired into your brand.
Improvement
Forgiving URL entry.
Forget the https://? No problem — Brail now reads a website address however you type it.
UX/UI
Activity at a glance on your overview.
Activity cards bring recent changes onto your overview, so you can see what's moving without digging.
UX/UI
A refreshed app shell.
A lighter top bar, a rounded search field, a refined logo and tidier spacing throughout — small touches that make the whole app feel sharper.
Improvement
A simpler brand model.
We streamlined your brand model — folding identity into governance, narrative into strategy, and moving governance alongside your other brand pillars. Fewer places to look, clearer structure.
Feature
SEO discovered during onboarding.
Onboarding now captures the search and discovery language your brand should own — so your agents write to be found, not just to be on-brand.
Improvement
Pull brand images from your site.
Onboarding now finds and pulls the best images from your website — giving your brand a head start on real assets.
Improvement
Live progress during onboarding.
A running progress panel shows Brail's work as it happens, so a thorough setup never feels stuck.
Improvement
Know the moment a proposal arrives.
A subtle indicator now flags new proposals as they come in, so nothing waiting for your review slips past.
Improvement
Smarter examples, with thumbnails.
Saving an example now generates a thumbnail and tightens how examples are created — a more visual, faster library.
Feature
Generate on-brand illustrations.
Brail can now produce illustrations in your visual style — on-brand imagery, generated on demand.
UX/UI
Edit your visual style directly.
Your visual pillar is now editable in place — adjust colours, type and style without leaving the page.
UX/UI
A lighter sign-in page.
A cleaner, brighter sign-in — a better first impression every time you log in.
Feature
An SEO language editor.
A dedicated editor for your search-and-discovery language — own the terms your audience actually searches for.
Improvement
Smoother file uploads.
A round of fixes makes uploading and ingesting files faster and more reliable, wherever you add them.
Feature
Pull audiences from your documents.
Upload a deck or doc and Brail extracts your audiences automatically — no retyping what you've already written down.
Feature
Fill the gaps in your brand.
A guided flow walks you through the parts of your brand that aren't complete yet — turning a sparse brand brain into a full one, one prompt at a time.
Improvement
Richer answers for your agents.
Connected agents can now pull image thumbnails alongside brand context — so what they generate is grounded in your real assets.
Feature
A library of on-brand examples.
A new home for the copy and creative that defines your brand — upload approved and rejected examples, including images, to teach every tool what good looks like.
UX/UI
Restructured navigation.
A clearer navigation structure that collapses detail by default — less to scan, faster to the thing you need.
Improvement
Better website crawling.
Smarter crawling pulls more of the right content from your site, with less noise.
Feature
A drag-and-drop brand overview.
A new overview page brings your whole brand into one place — rearrange the cards to suit how you work.
Improvement
No more disruptive auto-refresh.
We stopped pages from reloading underneath you — your place and your scroll stay put.
Connector
Connect Cursor to your brand.
Cursor now pulls your live brand context, so what you build there starts on-brand.
Connector
Connect Perplexity.
Bring your live brand context into Perplexity, so research and answers stay grounded in your brand.
Connector
Connect Microsoft Copilot.
Copilot now writes and builds with your brand context wired in — on-brand across your Microsoft tools.
Connector
Connect Claude to your brand.
Securely connect Claude to Brail and give it your live brand context — colours, voice, audiences and approved assets — as it works.
Improvement
Review every change before it goes live.
A refined proposal flow stages suggested changes for your approval — nothing touches your live brand until you say so.
UX/UI
A more polished onboarding.
Your logo now appears as you set up, with a smoother end-to-end flow — a warmer first run.
Feature
Guided brand setup.
A new onboarding wizard builds your brand with you — point it at your website and it drafts your strategy, voice and messaging for you to refine.
Improvement
Find your social profiles automatically.
Setup now detects your social handles from your site, so your channels are linked without the copy-paste.