Start the day with a brief, and images that know your brand

The Daily Forge Brief

A scheduled morning briefing that reads what happened overnight and turns it into a plan: what moved, trending conversations, competitor activity, opportunities that fit your brand, what to ignore, and anything needing caution. Alongside it comes a batch of up to 20 ready-to-review drafts across nine content types, from X posts and threads to reply packs, visual concepts and newsletter sections, with your own per-type targets.

• Every item is backed by linked sources with freshness labels, and a section with nothing solid to say stays empty instead of filling with filler
• Each draft carries a risk level and any policy warnings. Edit it and it is re-checked; save it, reject it with a reason, or approve it straight onto the content calendar with a platform and time slot
• Pick a timezone, a ready-by time and a daily or weekdays cadence. Teammates can be alerted in-app, by email or on Telegram, and alerts carry only a short summary and a secure link back into the app, never the draft content
• Owners and admins can generate a test brief at any time, before or after enabling it

The Daily Forge Brief is available to Web3 and Content Creator workspaces.

One-click Daily Brief setup

One click analyzes your workspace, from post history and engagement patterns to competitors, Brand Voice, Knowledge Base and calendar, and fills in the entire setup as a reviewable proposal. Nothing is saved or enabled until you confirm every field.

• Choose a planning horizon (this week, ongoing, or a custom range) and optionally state what the plan should build toward
• The proposal explains its recommendations and plainly flags where your workspace lacks the history to support one, instead of guessing
• A new Planning Focus section steers upcoming briefs toward a campaign objective, key messages, a call to action and constraints for a set date range, without permanently changing Brand Voice or content policy

Workspace strategy for one-off posts

A Use workspace strategy toggle in Content Studio carries your Daily Brief strategy into individual generations, with per-post overrides for audience, objective, angle and call-to-action posture. Restricted topics and approved-claim boundaries stay enforced no matter what you override.

• Workspaces that never set up a strategy get one inferred from their own published and approved content: typical length and shape, hook and call-to-action habits, topic pillars and audience. It is applied firmly only when there is enough history to be confident, and it never copies competitors
• Your toggle choice is remembered between sessions

A visual brand kit

Workspace branding settings now hold a full visual identity: whether the brand is a company, a creator or a mascot, logo and on-image text policies, motifs to prefer or avoid, and hand-picked reference images from your asset library, each tagged by what it teaches (identity, style, layout or subject).

• A readiness meter shows whether your kit is Generic, Guided or Locked, and says exactly what is missing
• Owners and admins approve individual references and templates; approved templates unlock the highest tier

Brand-aware images in Content Studio

Generated post images now follow your brand. Your colors, logo rules, style notes and approved references shape every picture, and a strategy picker lets you steer the look: editorial scene, illustration, mascot, product shot and more.

• Generate an alternate take on any image and flip between the two right on the preview
• Rate any image as on-brand or off-brand, or promote a great one into your brand reference library with one click
• If your kit is still empty, a one-time heads-up warns that results will look generic until you fill it in
• Reloading a saved post from history brings its images back with it

Generated images travel with your post

Publishing or scheduling from Studio now brings your selected images along automatically, with the right crop for each platform: widescreen for X and Facebook, square for Instagram. A preview card in the publish panel shows exactly what is attached where, and you can remove it or swap in your own media at any time.

• For threads, the image lands on the first tweet unless you attached per-tweet media
• The calendar shows image thumbnails on scheduled posts, and Studio history shows small previews of each saved post's images

Best mode is rebuilt around your brand

Best-quality generations in Rewrite Draft, Breaking News, Post Helper and Thread modes now compile a focused brief of your brand voice, approved facts, strategy and recent content, draft several genuinely different candidates, then check each for brand fit, freshness against your recent posts, length and factual grounding before picking the winner. This is on by default for every workspace.

• Results include a Why this sounds like your brand panel with plain-language reasons, the sources the post is grounded in, and review-before-publishing warnings when something needs a human look
• Thumbs up and down let you tell Studio whether a result sounds like you, and result badges show how many candidates were considered and the brand score
• The loading screen walks through the real stages (Context, Draft, Validate) with an honest 25 to 60 second estimate

Fast mode and Comment King are unchanged, and quote posts or posts referencing someone else's post keep the previous pipeline.

Posts tuned to current platform algorithms

Best-mode generations now read the newest daily algorithm brief for each platform, covering what is trending, what to do more of, what to avoid and format tips for X, Instagram and Facebook, applied as advisory guidance that never outranks your brand rules. Results report exactly which signals were applied, which were considered but skipped, and which platforms had a current brief.

Brand voice learns from better examples

Brand voice generation now removes duplicate posts, balances your best-performing examples with your most recent ones across every connected platform, and labels each example's engagement level, so a repetitive or one-platform-heavy history no longer skews the voice.

• A new one-click source, Use approved Daily Brief work, learns only from brief items you edited, approved, saved, scheduled or posted. Raw AI drafts are never training material
• A newly generated voice no longer silently replaces the one in production. It arrives as a clearly marked candidate, with a confirmation step before you activate it

Sharper generated graphics

Image generation moved to a new model. When your brand kit includes approved style references, it produces finished campaign graphics that match your brand's craft and picks a composition suited to the story: comparison, workflow, feature grid, data panel and more, with the headline rendered exactly once. The image control formerly labeled Template is now Composition.

Header images on the new model

Newsletter header images are generated by the new image model, and revisions now edit the existing image directly instead of regenerating from scratch.

Follow-up drafts that survive outages

If the primary AI writing service fails, follow-up drafts retry with a backup provider. If everything is down, you get a safe, simple fallback message built only from the contact's name and the property address, tailored to the sequence and step, so your follow-up sequences never stall.

• The fallback never invents dates, prices, features or market claims, and it goes through the same compliance, consent and governance checks as an AI-written draft

Compliance guardrails catch more risky phrasing

The checker now catches fair-housing and investment-promise wording that previously slipped through, such as family-targeted phrasing with words in between, and guaranteed-appreciation claims in their many forms.

Brand Voice results no longer crash the Studio

Newer brand voices store their tone rules and vocabulary in a richer format the results view could not display, crashing to a blank screen. Both formats now render, and a malformed entry is skipped instead of breaking the page.

No more duplicate results from long generations

A queued generation that took long enough to be delivered twice could leave two different history entries for one request, with the on-screen result flipping between versions. The first completed result is now kept and reused.

No more forced ALL-CAPS hooks

When your brand voice does not specify a hook style, posts no longer default to opening with an all-caps headline. Brands whose voice explicitly calls for one keep it.

Category renames save

Renaming a topic category in the category manager now actually saves; edits were being silently discarded.

No duplicate stories across categories

The same news story no longer appears under multiple categories in one daily run. Duplicates are matched by what the story is about, not just identical links, and the check holds even when categories generate in parallel. A category can now show fewer topics rather than a repeat.

Trend scan survives X rate limits

The trending-topics scan no longer fails outright when X's trends feed is rate-limited or empty. It falls back to a live search of X, so fresh trends still appear.

Honest status for Generate All

A run now reports success only when topics were actually saved. A run where everything failed is marked failed with the real reason, a partial result is flagged as partial, and a category stalled by a momentary hiccup is recorded as failed instead of stuck. The button stays in progress until the run truly finishes, and retrying after a network blip cannot launch the run twice.

Formatted listing description results

Generating a listing description now shows a clean result with separate Headline, Full description and Short teaser sections, each with its own copy button plus copy-all, and publishing is pre-filled with the headline and description.

Open-house content asks for the date first

You can no longer start open-house content without an open house date and time on the property. The generate button is disabled with a clear hint, and campaign items tell you to add the date instead of sending you into a generation that would fail.

Photo upload links create reliably

Creating an upload link for a seller or photographer reliably shows the new link, failures state the actual reason, and a previously created link is cleared from view when you start a new one so you cannot copy a stale link.

Portal share links work again

Sharing a property portal with a client works again; the secure share token is now created on the server when the share is made.

Workspace switching is reliable

Switching workspaces now lands you where you picked and keeps you there. The switch no longer triggers a surprise reload midway, bounces you back after the page loads, or quietly snaps back a moment after it looked done. Switching back from Investor Intel to a content workspace is covered too, and a switch that fails partway still returns you safely to where you were.

Social sync fills the gap

Syncing a connected social account now pulls everything since the last successful sync, up to a year back, instead of only the last 24 hours. Posts published during an outage or a pause no longer silently go missing from analytics and history.

One record per social post

A post imported from a CSV and also pulled in by sync no longer counts twice in analytics or appears twice on the calendar. Sync now recognizes a post regardless of which path brought it in, and keeps calendar entries up to date with the post's real text and publish time.

The Extensions page loads from anywhere

Opening the browser Extensions page directly, by refreshing, bookmarking or following a link, showed page not found. It now loads correctly every way you reach it.