# Stu AI - Your Content Studio & Team: every page in full > An AI content engine that drafts campaigns toward a goal you set and counts results only with evidence. Nothing is scheduled or published until a person approves it. Studio is made by Bolls in Nashville, Tennessee. What this site does not claim: it names no customers, and it names no logos, testimonials or metrics. It publishes no citation numbers. # Home Source: https://stu.now/ ## Lead your category. Set the goal. Studio drafts the work. Nothing goes out until you approve it. - You set the goal - Every run on the record - Each piece approved on its own New: Every agent run is on the record, with what it did and the model spend behind it. ## You hand out the work. It comes back checked. Pick a step and watch the screen change. Each one is real Studio UI, drawn in code and filled with a made-up workspace, carrying the words the product itself uses. ### 01. Plan: You say it once. It comes back as a plan. One field, not a chat window. Studio restates what you asked for as a name, an outcome, an audience, timing, a first plan and a goal, and then waits. Nothing is created until you start it. Where: Campaigns, the start flow ### 02. Research: Sources, and whether the page really says it. The sources behind your articles, and whether each page really states the figure it is cited for. Verify sources from the Sources screen, and a figure that is not on its page is flagged for a person to open. Where: A campaign's sources ### 03. Draft: A draft arrives with its sources attached. Drafted from your own guides, not a blank prompt. Every figure carries the number of the source it came from, and the source stays on the page through the check and through your approval. Where: The draft, at Publish ### 04. Checks: Four checks, and the one that failed. The four gates an approval records, with every standard under them. A check that fails is not a warning you click past: the approve button becomes the refusal and names the count. Where: The draft, at Checks ### 05. Approve: Nothing goes out until you say so. The gates as the record will store them, the byline, and the choice of what approving does: keep an approved draft, set a date, or let it go out. Today each piece is its own decision, and nothing moves while you have not made it. Where: The draft, at Approval ### 06. Results: Counted only with evidence. A goal with a number and a date, a window that opens at each publish, and results that exist only because something can show them. Anything a person types in is labeled attested. Without evidence the campaign says so, and says how to add some. Where: A campaign's own page, at Outcomes ### What the Research step shows Sources for this campaign: Checked for complete fields, and against what each page states. Each source is listed with its publisher, how the draft used it, and a pass, a warning or a fail. A source missing a field fails and blocks the approval ("No URL on the source"); a figure the page was read for and not found on warns and asks a person to open it ("Figure not found on the page"). ## Every draft is checked, and names what needs you. Open a draft in Review and Studio runs its checks: voice rules, the sources behind each figure, and your product facts. Each is a pass, a warning or a fail, and a fail keeps the piece on your desk until it is fixed. The marks on this draft illustrate those results. ### Your voice, then a voice check Studio writes from your brand voice guide, then checks the draft for the things good brand writing avoids: hype phrases, sales calls to action, exclamation marks and too many dashes. These are warnings for you to weigh, not blocks. ### Every figure, against the page it came from Each figure should cite a page that states it. When Studio has read the page and cannot find the figure, the check says which one, so a person can open the page and fix the figure or the source. ### Claims, against your product facts Claims are checked against your product facts, and a claim that does not match them is flagged for you rather than quietly left in. ### A fail you cannot approve past Some checks are structural: a source missing its URL, or a footnote with nothing behind it. Those fail, and the piece stays with you until they are fixed. A warning is the other kind, and you can approve past one. | Check | What it looks at | Result it can record | | --- | --- | --- | | Brand voice | Generic voice rules: hype phrases, sales CTAs, exclamation marks, dash budget | pass or warn | | Sources / citations | Complete source fields and footnote mapping, and whether a page states the figure it is cited for | pass, warn or fail; only a structural problem fails | | Product canon | Whether a draft's claims match the workspace's product facts | pass or warn | | AEO | Whether an answer engine can quote it: a direct answer, questions as headings, FAQs, a sourced statistic, structured data | pass or warn | All four are recorded with the approval. A warning is yours to weigh. A failed check has to be fixed before this can be approved. - Brand voice (pass): No hype, no sales CTAs. The voice rules are generic ones: hype phrases, sales calls to action, exclamation marks and the dash budget. - Sources / citations (pass): Footnote 2 points at a listed source. Every footnote in this draft maps to one of its four sources. - Sources / citations (warn): This page may not state this figure. Studio could not find "about a third" on Mid-Market Benchmark's survey page. Open the page, then fix the figure, change the source, or mark it checked. - Sources / citations (fail): Sources have title, publisher, and URL. Incomplete source: #3 (missing url). Every source needs title, publisher, and URL. The results above are the product's. Studio shows them as a list in Review and does not mark text inline; the sample draft on the how-it-works page places each one beside the words it is about. A draft waiting on a person reads "Waiting on your approval". Nothing happens to this piece until you decide. ## A campaign never reads On track without evidence. Not from a publish, and not from a count of zero: only from outcomes it can show you. An outcome a person types in is always labeled attested. ## It researches and drafts. You decide what goes out. Each run says when it ran and why and what it wrote, and every model call it makes is metered to it, in tokens and in dollars. ### What is true today Agents research and draft on their schedules. Nothing they make goes out without a person. Claim changes are coming later, as an opt-in. - Research: On its schedule - Drafting: On its schedule - Claim changes: Coming later (opt-in) - Publishing: Approval required ## Each item is approved on its own. An article goes out only after a person approves that article. Coming for partners: if your site is built from a GitHub repository, Studio opens a pull request for each approved article. Early partners receive approved drafts to publish themselves. - Your site (Coming for partners): If your site is built from a GitHub repository, Studio opens a pull request for each approved article. - Buffer (Coming for partners): If you connect your Buffer account, a post goes through it once you approve that post. Early partners receive an approved draft of each post for LinkedIn, X, Facebook and Instagram. ## In your voice, with its receipts. - Your voice: Studio writes from your own guides, not from a blank prompt. - Its sources: Every source it cites stays on record, with a check of whether the page really says what it is credited with. - Your reasons: When you turn something down, say why. Your reasons become a memory you can read, and later work follows it. ## Written to be quoted by AI answers. Before you approve an article, Studio checks it the way answer engines read it. - A direct answer - Headings phrased as questions - A short answer under each one - A set of FAQs - At least one sourced statistic - Structured data that will validate Coming to design partners. Studio tracks AI citations through Profound and Promptwatch. - Profound - Promptwatch ## Security - Membership is checked: Every request checks that the person signed in is a member of the workspace. - Credentials stay sealed: The publishing tokens and model keys you connect are encrypted at rest, used only on the server, and shown back as their last four characters. - Roles decide who approves: Only someone with the approver role can approve an article, and agents connected over MCP get no key that can approve, schedule or publish. - Spending has a ceiling: Each workspace has a monthly model budget. When it is spent, the work stops and says so, instead of running up a bill. ## Plans are priced. Signing up is free. Run one real campaign with us. You approve every draft. No card needed. ## FAQ ### What does Studio make? Articles for your blog, posts for your social channels and the briefs behind them, organized into campaigns that each pursue one goal. Agents research and draft on their schedules. Nothing they make goes out without a person. Studio also watches your sources and the news for openings worth answering, and files them as ideas. ### Where does approved work go? Coming for partners. If your site is built from a GitHub repository, Studio opens a pull request for each approved article, and if you connect your Buffer account, a post goes through it once you approve that post. Early partners receive approved drafts to publish themselves. Each article and each post is approved on its own. ### Does anything go out without a person? Not today. You stay in control: nothing is scheduled or published until a person approves it, and each piece is approved on its own. Articles need someone with the approver role. Posts wait for a person too, and every agent run is recorded with what it did. Rules you trust that approve for you are coming later, as an opt-in. ### How is a goal measured? A goal names a metric, a target, a deadline and a lookback of 30, 60 or 90 days. A result counts for every piece published within the lookback before it, until the deadline's day ends, and its credit is split across those pieces: each one's weight halves for every seven days between its publication and the result, and the shares add up to the whole result. A campaign says On track only when outcomes back it. ### Can it tell us whether AI answers cite us? Coming to design partners. Studio tracks AI citations through Profound and Promptwatch. Every article is also checked before approval for what answer engines look for: a direct answer, questions as headings, FAQs, a sourced statistic and structured data. ### Can our own agents work with it? Yes. An agent you run elsewhere can connect over MCP with a key you scope. It can file ideas and content plans, read your campaigns and an article's review, and record outcomes, which are always labeled attested. There is no key that can approve, generate, schedule or publish: your agents propose, and a person decides. ### What does it cost? Free stays free. Starter is $149 a month, Growth is $399, Scale is $999. Enterprise is custom: talk to us. There is no checkout yet, so every plan still starts as a conversation. ## Set the goal. Studio drafts the work. Nothing goes out until you approve it. # How it works Source: https://stu.now/how-it-works ## From a sentence to a campaign that counts. Studio takes what you want a campaign to achieve, does the work in the open, waits for your decisions, and keeps the score against a target you set. ## 01. Say it: Say what the campaign should accomplish. Type it the way you would say it to a colleague. Studio restates it as a campaign for you to confirm: a name, the outcome, the audience, the timing, the pillar it belongs to, a first plan and a goal. - Nothing is created until you start it, and pressing Start twice still starts one campaign. - Change something before you start, or cancel, and nothing is left behind. - Start from a typed sentence, from a Radar opportunity, or by growing an existing article or post into a campaign. Every start goes through the same restatement. ## 02. It works: It does the work, on the record. Agents research and draft on their schedules. Studio watches your sources and the news for openings worth answering, drafts articles and posts in your voice with their sources, and proposes dates for the work. Nothing they make goes out without a person. - Agents you set up on News Radar watch the topics and pillars you choose, and each one's work is recorded run by run. - Drafts follow your own guides. Channel and persona voices are coming. - Every source a draft cites stays on record, with a check of whether the page really says what it is credited with. ## 03. You decide: You decide what goes out. A campaign's page shows what waits for you: drafts in review, posts awaiting approval and recommendations to decide, the most urgent first. - An article goes out only after someone with the approver role approves it on its Approval tab, and the approver can keep it as a draft. - Coming for partners: if your site is built from a GitHub repository, Studio opens a pull request for it, and a post goes through Buffer once that post is approved, if you connect your account. Early partners receive approved drafts to publish themselves. - Each article is approved on its own. Each post is approved on its own, by a person. - A recommendation's changes apply once. If the work moved on since, the change is not applied and the card says why, and published work is never rewritten behind your back. ## 04. It counts: It counts what happened, against your goal. A goal names a metric, a target, a deadline and a lookback of 30, 60 or 90 days. Publishing opens the window, and the campaign reads Measuring while it runs. - A result counts when it lands within the lookback after one of the campaign's pieces is published, until the deadline's day ends. - A result's credit is split across the pieces it counts for. Each piece is weighted by how long before the result it was published, halving for every seven days, and the shares add up to the whole result. A piece published more than once counts from its publication closest before the result. There is no data-driven attribution. - On track and Complete come only from evidence. Without it, a campaign says so and tells you how to add some: connect a source, or attest an outcome, which is always labeled attested. ## What a campaign tells you One state leads, and everything else that needs a person stays listed beneath it. | State | What it means | | --- | --- | | Needs attention | Something failed to publish. | | Needs judgment | An agent held a post for a person. | | At risk | The month's model budget is spent while work is unfinished. | | Waiting on approval | An article or posts are waiting for review. | | Measuring | Its work is published and the goal's window is open. | | On track | Outcomes back it, and it says which. | | Complete | The window closed, with evidence. | | Ended, not measured | The window closed without evidence, and it says so. | ## What every run records. Each time an agent works, the run is written down, whether it was started on a schedule, by a person, or by your own agent over MCP. - When, and why: Its start, how long it took, and what started it: its schedule, a person, or MCP. - What it wrote: The opportunities, promotion runs and campaign work it produced, each linked back to the run. - What it spent: Each model call it made, metered to the run and the agent, with its tokens and its dollars. - The Boundaries it read: The workspace's Boundaries as the run read them when it started. Agents research and draft on their schedules, and nothing they make goes out without a person. ## Publishing waits for a person today. Agents research and draft on their schedules. Nothing they make goes out without a person. Claim changes are coming later, as an opt-in. # Security Source: https://stu.now/security ## Built for the person who can say no. How Studio checks membership, what it seals, what it leaves to people, and what it caps and writes down. Each point comes from the product's code; none is a certification. ## Membership is checked Every request checks that the person signed in is a member of the workspace. - Members sign in with a link sent to their work email. That sign-in link works once and expires after 15 minutes. ## Credentials stay sealed The publishing tokens and model keys you connect are encrypted at rest, used only on the server, and shown back as their last four characters. - The tokens, app passwords and API keys you connect are encrypted at rest with AES-256-GCM under a server-side secret, and decrypted only on the server, at the moment they are used. - What a screen reads back is whether a key is set, its last four characters, and when it changed. The key itself is not part of that, so it does not reach the browser. - Destination settings accept only their own named fields, and tokens are stored only in the encrypted store. - Keys for outside agents are stored hashed, so a copy of the database is not a set of working keys. ## Roles decide who approves Only someone with the approver role can approve an article, and agents connected over MCP get no key that can approve, schedule or publish. - An article goes out only after someone with the approver role approves it, and a failing check has to be resolved first. - Agents research and draft. They cannot schedule or publish until a person approves. Only an admin sets what agents may do. - Agents you run elsewhere connect over MCP with a key you scope. No scope can approve, generate, schedule or publish, and a revoked key keeps its name on everything it did. ## Spending has a ceiling Each workspace has a monthly model budget. When it is spent, the work stops and says so, instead of running up a bill. - Every model call is metered to its workspace, and a call an agent makes is also tied to its run and its agent. - When the month's budget is spent, Studio stops writing, and each campaign with unfinished work reads At risk and says that is why. ## Everything is on the record Runs, decisions and outcomes are written down as they happen, and a correction is a new row rather than an edit. - Every agent run records when and why it ran and what it wrote, and each model call it makes is metered to that run. - A decision's changes apply once. A change is applied only against the value it was formed from, so one whose work has moved on is recorded as not applied instead of overwriting the newer work. - Outcomes form an append-only ledger. A correction is a new row, an attested outcome is always labeled attested, and only an owner or an admin can redact an attested statement, which is itself recorded. This page makes no compliance claim: no certification, attestation or audit is described here. # Sign in Source: https://stu.now/signin ## Sign in to Studio Sign-in is by email. Enter your work address on the next page and we'll send you a link that signs you in. It works once, and expires after 15 minutes. No workspace yet? Sign up and create one, or talk it through with us first. Sign up: https://studio.stu.now/studio/onboard. Talk to us: hello@stu.now.