¿Cómo podemos ayudarte?

STETLY is the check between draft and publish. You teach it your brand once (voice, style rules, facts, examples), and every draft you or your team writes gets checked against that brand before it goes out.

Every check returns two scores:

• Brand Integrity (0 to 100) measures how well the draft matches your voice, style guide, and brand facts.
• SEO Visibility (0 to 100) measures how discoverable the content is. It appears for SEO-relevant content types such as website copy and blog posts.

Beyond checking, STETLY can create content that follows your brand (Compose), answer questions about your brand (the Stet assistant), and enforce contract obligations such as sponsor or client requirements (Brand Insurance).

The name comes from the proofreading mark "stet", Latin for "let it stand". Every check ends with your voice left standing.

1. Go to the app and open the Register tab.
2. Enter your name, company name, email, and a password, then accept the terms. Your company name pre-fills your first workspace name.
3. Confirm your email through the link we send you.

You can also sign in with Google, or use a magic link: we email you a sign-in link and you skip the password entirely.

While demand is high we onboard every customer personally. After signing up you'll see a welcome screen where you book a short onboarding call; your workspace goes live right after the call. It's how we make sure your brand is set up properly from day one.

Your first workspace starts with a 7-day free trial with full access. No credit card required.

A workspace is your billing and team boundary. It has a plan, a monthly credit allowance, and members. Agencies typically use one workspace for the whole team.

A brand lives inside a workspace. Each brand has its own voice profile, style guide, knowledge library, and check history, fully isolated from other brands. An agency managing five clients creates five brands.

Switch between workspaces and brands with the switchers in the top bar. Most screens in the left navigation (checks, Compose, style guide, content) act on the brand you have selected.

1. After creating your workspace you land on the Dashboard. Choose Create the first brand.
2. Give the brand a name. Voice summary, target audience, and tone words are optional; the style guide wizard can fill them in for you.
3. Pick the brand's default language.
4. You land on the Style Guide page. Choose Generate to let STETLY build your style guide from your existing content (recommended), or Create manually to write rules yourself.

For the full wizard walkthrough, see "Building your style guide automatically" in this help center.

While your brand is new, the Dashboard shows a short checklist: create a brand, add content, build your style guide, run your first check.

It tracks progress automatically and disappears once you've completed everything.

1. Open Run a Check in the left navigation.
2. Pick a content type, for example Blog post, Email, or Instagram post. The style guide wizard created these for you, and you can add more later.
3. Give the submission a title, paste your draft (or fill in the fields for multi-field types like Email), and press Run check.
4. Watch the progress steps, then review your scores and findings.

The Running checks section of this help center covers everything the results screen can do, starting with "Running a check, step by step".

Every term you'll meet in the app, in one place.

• Check: one evaluation of a draft against your brand's rules.
• Submission: a tracked piece of content with its version and check history.
• Brand Integrity score: 0 to 100 measure of how on-brand a draft is.
• SEO Visibility score: 0 to 100 measure of search discoverability, shown for SEO-relevant content types.
• Pass threshold: the score a check must reach to pass, set per brand.
• Blocking rule: a guideline that fails a check when violated, regardless of score.
• Guideline: a single enforceable brand rule.
• Style guide: all of a brand's guidelines, voice profile, and facts together.
• Content type: a category of content with its own rules and fields (Email, Blog post, Instagram post).
• Compose: STETLY's on-brand content generation.
• Draft: a piece of content generated by Compose.
• Knowledge library: everything ingested about your brand (documents, posts, pages, images).
• Fact: a concrete brand statement enforced during checks.
• Suggestion: a rule or fact STETLY proposes for your review before it becomes active. Also called a proposal.
• Brand Insurance: contract-derived compliance rules enforced as blocking guidelines.
• Stet: the brand-aware assistant.
• Credits: the single usage meter for all AI actions.
• Workspace: billing and team boundary; holds brands and members.
• Brand: one brand's isolated profile, rules, content, and history.
• BYOK: bring your own key; use your own AI provider keys on the Business plan.

1. Open Run a Check.
2. Pick a content type. This is required: it determines which rules apply, and each content type has its own checklist derived from your guidelines.
3. Pick a subcategory if the type has more than one. It narrows the rule set further, for example separate rules for promotional and educational Instagram posts.
4. Give the submission a title. This is how the check appears in Past Checks.
5. Optionally pick a channel (Instagram, X, LinkedIn, or Email) to get a platform-styled preview on the result.
6. Add your content. Paste text, or fill in the type's fields (see "Multi-field content types" in this help center). For the Website content type you can switch to URL mode, paste a link, and press Fetch to pull in the page text automatically.
7. Press Run check. Checks run in the background, so you can watch the progress steps or leave the page and come back.

Each check uses credits from your workspace allowance. See "What costs credits" in this help center.

Some content types are made of several fields that get checked together as one piece. An Email has a subject line and a body. An Instagram post has a visual (an image upload) and a caption.

You fill in each field separately, and the check evaluates the whole piece in one pass. Every finding is tagged with the field it belongs to, so you know whether the problem sits in your subject line or your body copy.

Image fields accept uploads and are evaluated visually against your visual guidelines.

Field layouts are defined per content type. The style guide wizard sets sensible defaults, and you can adjust them per brand.

Brand Integrity (0 to 100) is the weighted result across all criteria in the checklist. Higher-weighted guidelines move it more.

SEO Visibility (0 to 100) is a separate score built from SEO criteria. It only appears for content types where SEO is relevant, such as website copy and blog posts. An Instagram caption is never penalized for missing keywords.

Your brand has a pass threshold (default 70, editable in Brand Settings). A check passes when the score clears the threshold and no blocking rule was violated. See "Blocking vs warning rules and the compliance gate" in this help center.

The Insights dashboard shows your score history over time.

Findings are grouped by severity: errors, warnings, and info. Each finding shows:

• The guideline it violated.
• The exact evidence, meaning the quoted passage from your draft where the problem occurs.
• A suggested rewrite where possible, with the original text, the proposed replacement, and a copy button.

Use the severity tabs to triage. Fix errors first, review warnings, skim info.

Every guideline has a severity.

A blocking rule fails the check outright when violated, regardless of the numeric score. Use it for hard rules such as legal or contract obligations: "never mention competitor X", "always include the ad disclosure".

A warning rule counts against the score but never blocks on its own.

This is the compliance gate: a draft with a great score can still fail if it breaks a blocking rule. That's intentional. It's the safety net that lets STETLY enforce your contracts as Brand Insurance.

Findings that can be repaired come with a suggested rewrite inline. You can copy the suggestion and apply it in your own editor, or press Fix this to have STETLY apply the edit for you and produce an updated version of the submission.

Fixes are targeted. Only the flagged passage changes; the rest of your draft is left standing.

After editing a draft, open the submission and press Recheck. This runs a fresh check and adds a new version to the submission, so you keep the full history.

• The version picker on the submission page lets you view any earlier version and its results.
• Score deltas show how much each revision improved.

Rechecks run in the background like normal checks and cost credits.

Past Checks lists every submission for the current brand with its status, content type, latest score, and date. Filter by approval status, content type, or date range.

Click any submission to open its full result, version history, and actions.

For teams that want an explicit sign-off step, every submission carries an approval status. It moves from Draft to Pending review, then to Approved or Rejected.

• Reviewers open the submission and press Approve or Reject. A rejection includes a reason, shown as a banner on the submission.
• The submission page shows the approval chain and each reviewer's status.
• Past Checks can be filtered by approval status, so approvers see their queue at a glance.

STETLY runs a quick evaluability pass before the full check. If the submitted content is a placeholder rather than a real draft (just "Test", or lorem ipsum), the check stops early instead of producing a meaningless score, and the credit is refunded.

Submit real draft content to get a real result.

Content types are the categories your content falls into: Email, Blog post, Instagram post, Website copy. Each type has its own rules, its own checklist, and optionally its own fields (an Email has a subject and a body, an Instagram post has a visual and a caption).

Most types come from the style guide wizard, which creates the ones it finds in your content. You can add your own at any time for formats the wizard didn't cover, such as a press release or a product description.

The fastest way to create one is to ask Stet. Open the chat and describe what you need, for example: "Create a content type called Press release with a headline and a body." Stet asks for whatever is missing, shows you a summary, and creates the type once you confirm. The matching checklist is set up automatically, so the new type is immediately available in Run a Check and Compose.

When you create a type, you decide three things:

• Its name, for example "Press release".
• Its fields, when the format has separate parts that should be checked together (subject + body, visual + caption). Skip fields for plain single-text formats.
• Whether it is SEO relevant. SEO-relevant types such as website copy and blog posts get the SEO Visibility score alongside Brand Integrity.

A new type starts with no rules of its own, so add guidelines scoped to it; see "Brand guidelines vs content guidelines". You can ask Stet for that too: "Add a guideline for press releases: keep the headline under 80 characters." Brand-wide guidelines apply to every type automatically.

BrandVoice Certified is a public badge that shows your brand consistently writes on-brand content, verified by your own check history rather than a one-time audit.

To earn it, a brand needs: a compiled brand voice with at least one active guideline, at least 8 checks run in the trailing 30 days, and an average Brand Integrity score of 70% or higher across those checks. STETLY re-evaluates every brand daily, so certification tracks your recent activity, not a snapshot from months ago.

If your average score or check volume dips below the bar, the brand has a 7-day grace period before certification lapses — a single off day won't cost you the badge, but a sustained drop will.

Certification is available on plans that include it (Pro and Business; see /pricing). Once eligible, open Brand Settings → BrandVoice Certified to publish: turning on Publish public certification page gives you a public verification page and an embeddable badge, plus the option to show your live score alongside the status. Copy the embed snippet to add the badge to your own site.

The badge updates hourly once cached, so a status change (earning or lapsing certification) can take up to an hour to show up everywhere it's embedded.

The style guide wizard reads your existing content and drafts your whole brand setup: voice profile, guidelines, content types, and brand facts.

1. Open Style Guide and press Generate (or Update guidelines if you already have a guide).
2. Choose your content source: connect Instagram, pick a Google Drive folder, upload files, paste text, or add URLs. You can combine sources.
3. Choose what to generate: brand voice, brand guidelines, per-content-type guidelines, brand facts, or any combination. Regenerating one scope leaves the others untouched.
4. Choose the level of detail. Concise gives you fewer, sharper rules. Balanced, the default, covers more ground.
5. STETLY reads the content and drafts everything. This can take a few minutes. You can close the wizard and let it continue in the background; the Dashboard shows a resume banner while it runs.
6. Review the draft: voice summary, tone words, guidelines, and facts. Accept to publish, or edit first.

A few things to know:

• The wizard writes in your workspace's default language. You can change the language per run.
• If the draft would exceed your guideline budget, STETLY shows a curation preview that merges near-duplicates before saving. See the article on guideline budgets.
• Each plan includes a number of wizard runs per month; the wizard shows how many you have left. Failed runs don't count against you and can be retried with the same inputs.
• A full style guide build costs about 310 credits.

Brand guidelines apply to everything the brand publishes: voice, terminology, formatting rules.

Content guidelines are scoped to one content type, for example rules that only apply to Instagram captions or emails. They can be narrowed further to a subcategory.

Both live under Style Guide in the left navigation, each on its own page. At check time, both feed the checklist for the content type you're checking.

1. Open Brand Guidelines or Content Guidelines.
2. Press Add manual guideline.
3. Write the rule, pick its category and severity (blocking or warning), and save.

You can edit or delete any guideline inline, including ones the wizard generated. Deleted guidelines are archived rather than erased, so your past check results still make sense.

When STETLY compiles new content, after you ingest documents, sync Instagram, or extract a contract, it never changes your style guide on its own. It creates suggestions that wait for your review on the Style Guide page.

• Accept turns a suggestion into a live guideline or fact.
• Reject dismisses it. Rejected and deleted items are remembered and won't be re-proposed.

Compliance suggestions can be edited before you accept them, for example to adjust the severity or expiry date.

Nothing becomes active without your approval.

Guidelines pile up over time and start to overlap. The redundancy audit finds those overlaps and proposes merges.

Start it from the button on the guidelines pages. It scans your set in the background, clusters near-duplicate rules, and proposes a merge for each cluster. A merge keeps the coverage of both rules in a single cleaner rule.

Review and apply merges from the results panel. On large sets the audit can take a few minutes; you'll see progress and can leave the page while it runs.

Each scope has a cap on how many guidelines it can hold: brand-wide, per content type, and per subcategory. Budgets keep checks fast, focused, and affordable.

When a save would push a scope over its budget, STETLY proposes a consolidation plan instead of silently cutting rules. The plan merges or drops the weakest overlapping rules, and you review it before anything changes.

Blocking rules are never trimmed automatically.

Every guideline has two settings that control how it behaves in a check:

• Severity: a blocking rule fails the check when violated, regardless of the score. A warning rule only affects the score.
• Weight: how much the rule moves the Brand Integrity score relative to other rules.

The wizard sets sensible defaults, and both are editable per guideline.

For how blocking rules interact with the pass threshold, see the article on blocking vs warning rules and the compliance gate.

Compose is a conversation. Open Compose under Content Creation and tell Stet what you want to create, in plain language:

• "Draft an IG post about our June launch"
• "Based on my last 10 posts, write a follow-up"
• "Write a launch email for the new pricing"

There's no form and no content type picker. Stet works out the format from your request, asks a short question if something essential is missing, and the draft appears in the panel next to the chat (on smaller screens, tap View draft). Stet can also pull from your past content first, for example your recent posts, so follow-ups actually follow up.

Every draft is built from your brand profile: voice, tone words, guidelines, and facts. It follows the same rules your checks enforce.

The conversation is the workspace. Keep talking to change the draft, and everything stays in one thread; see "Revising a draft".

Drafting costs credits; see "What Compose costs".

Two ways to change a draft, and you can mix them freely:

• Ask Stet in the chat. "Make it punchier", "shorter subject line", "add a call to action". Each revision updates the draft in the panel, and the conversation doubles as your revision history. Want options? Ask for them: "give me two alternative angles" works the way the old variations button did.
• Edit the fields directly. Every text field in the draft panel is editable inline, with a character counter where the format has a length hint. Press Save to keep your edits.

Image fields show a visual brief (what the picture should show) with a copy button, so you can hand it to whoever creates the visual.

If the draft was generated with assumptions ("assumes it's for LinkedIn"), they're listed above the fields so you can correct them in chat instead of discovering them after publishing.

Press Check this in the draft panel to run the draft through your brand checks, exactly like pasted content.

The check runs in the background and takes you to the full result: scores, findings, and version history, the same view every submission gets. From there you can fix findings and recheck, or go back to the Compose conversation and ask Stet to fix them for you.

You can also make checking part of the conversation: tell Stet "and make sure it passes my brand checks" and the draft comes back checked and self-corrected. That uses more credits than a plain draft.

Every draft you compose is saved automatically, together with its conversation.

Past Drafts (under Content Creation) lists everything you've composed for the current brand. Opening a draft brings back more than the text: the original Stet conversation is restored too, so you can pick up exactly where you left off, "make the intro warmer" included.

Delete drafts you no longer need from the same list.

Composing uses credits from your workspace allowance:

• A new draft costs a compose action.
• Each revision you ask for in chat costs a smaller refine action.
• Asking Stet to check and self-correct the draft ("make sure it passes my checks") costs more than a plain draft, because it runs your brand checks on top.

Editing fields by hand and saving is free. Costs vary with content length; the Billing page shows exactly where your credits went this period. For the bigger picture, see the article on what costs credits.

Everything STETLY knows about your brand comes from the content you feed it: past posts, website pages, documents, product sheets.

That content powers the style guide wizard, the compiled brand profile used in checks and Compose, and Stet's answers.

The Knowledge page shows everything that has been ingested. Filter by source (Instagram, Google Drive, uploads, URLs, pasted text) or search by text.

Open Add Content and pick a source:

• Upload files: PDF, Word, Markdown, or plain text. Drag and drop or use the file picker.
• Paste text directly.
• Extract from a website: paste a URL and STETLY fetches the page text, with a preview before saving.
• Instagram: connect the account and your posts, stories, and reels sync on a schedule.
• Google Drive: connect and pick a folder; documents in it sync automatically.
• Notion: connect with an integration token and the pages you share sync in.

Connection details for Instagram, Google Drive, and Notion are in the integrations articles.

Stored content counts against your workspace's storage allowance.

New content is processed in the background: text is extracted, images are captioned, and everything becomes searchable.

STETLY then compiles what it learned into suggestions. New facts, guideline proposals, and exemplars appear in the review queue on the Style Guide page.

Your live style guide doesn't change until you accept a suggestion. See the article on review suggestions for how the queue works.

Facts are concrete statements about your brand: "founded in 2019", "we never discount", "the product name is always spelled lowercase". They come from the wizard, from compiled content, or you add them manually.

Facts are enforced during checks, so a draft that contradicts one gets flagged. Compose and Stet use them too.

Deleting a fact marks it superseded. STETLY remembers the deletion and won't re-propose the fact from the same sources.

Delete individual items or whole documents from the Knowledge page.

Removing a document also detaches whatever was derived from it. Facts and guidelines you already accepted stay in place until you remove them yourself.

Sponsor agreements, client contracts, and legal requirements contain obligations your content has to respect: mandatory disclosures, exclusivity clauses, forbidden claims.

Brand Insurance turns those documents into enforceable check rules, so a draft that breaks a contract can't slip through.

Brand Insurance is available on plans that include it; see /pricing for the current plans.

1. Open Add Contracts.
2. Upload the document (PDF, Word, Markdown, text, or HTML) and fill in the title, counterparty, and effective and end dates.
3. Press Upload and extract. STETLY reads the contract and extracts its content obligations in the background.
4. Extracted obligations appear as compliance suggestions in the review queue. Review each one, adjust the severity or expiry if needed, and accept.

Accepted obligations become blocking guidelines by default: a draft that violates one fails the check regardless of its score.

Compliance lists every active obligation, grouped by source contract. For each obligation you see:

• The obligation text and how many checks it has been applied in.
• An expiry badge when it has an end date: green means more than 30 days left, amber means it expires within 30 days, red means expired.
• A link to the evidence in the source contract.

Expired obligations stop being enforced, so a sponsorship that ends doesn't keep blocking your content.

Stet is the assistant built into STETLY. You'll meet it in two places: the Chat page, and the Compose page, where the whole drafting flow is a conversation with Stet.

It's not a generic chatbot. Stet is grounded in your brand: it can search your knowledge library, read your guidelines and brand facts, look through your check history, and draft content in your voice.

It also knows STETLY itself. Ask it how something works and it answers from the same help articles you're reading now, and can take you to the right screen.

Things people actually ask it:

• "What does our style guide say about emoji?"
• "Summarize why last week's newsletter failed its check."
• "Draft three subject lines for the spring campaign."
• "Where can I change my pass threshold?"
• "Create a content type for press releases."

For the setup side, see "Letting Stet set things up for you".

Conversations with Stet are organized into threads. Start a new thread per topic from the sidebar so each conversation stays focused.

Responses stream in live, and Stet shows which documents and guidelines it consulted, so you can see where an answer came from.

When Stet proposes anything that would change your workspace, a confirmation card appears in the chat with the exact change spelled out and Confirm and Reject buttons. Nothing runs without your go-ahead, and the card shows "Done" once the change is applied.

Stet is aware of the screen you're on. If you're viewing a submission and ask "why did this fail?", it knows which one you mean.

Stet messages use credits, like other AI actions in STETLY.

You don't have to click through settings to configure STETLY. Describe what you want in the chat and Stet sets it up, from any screen. It can:

• Create a content type, including multi-field layouts: "Create a content type called Press release with a headline and a body."
• Add a guideline, brand-wide or for one content type: "Add a blocking rule: never mention competitor pricing."
• Add a brand fact: "Remember that we were founded in 2019 in Berlin."
• Change brand settings like the pass threshold, tone words, audience, or keywords: "Raise our pass threshold to 80."

If something's missing, Stet asks before acting: for a new content type it might ask whether the format needs separate fields or whether it's SEO relevant. Once it has everything, you get a confirmation card describing exactly what will be created or changed. Nothing happens until you press Confirm.

The same works in the other direction. Ask "where can I invite teammates?" and Stet explains the steps and offers to take you to the screen, or simply to handle it in chat when it can.

Insights shows how your brand is doing over time. Pick a time range at the top: 7, 30, or 90 days, or all time.

The at risk strip at the top surfaces anything trending the wrong way, so you see problems before you go looking for them.

Below that, the dashboard covers pass rates, score trends, and which rules your team struggles with most. Each metric is explained in the article "Every Insights KPI, explained".

Here's what each KPI on the Insights dashboard measures.

• First pass rate: share of submissions that passed on their first check. The single best "are we on brand?" number.
• Rewrite lift: average score improvement between a submission's first and latest version. Shows how much the fix loop is worth.
• Time to approval: average time from first check to approval.
• Score over time: your Brand Integrity trend (and SEO Visibility where it applies), with your pass threshold drawn in.
• Most failed rules: the guidelines your team violates most. Useful for training, or for spotting a rule that's too strict.
• Score by content type: average score and pass rate per content type, so you can see which formats struggle.
• Checks per period: check volume over time.
• Rising / falling rules: guidelines whose pass rate is improving or declining.
• Evidence coverage: share of guidelines backed by evidence from your actual content.

Workspaces (in the workspace switcher) lists every workspace you belong to.

Create Workspace starts a fresh one with its own plan, credits, and members. The free trial applies to your first workspace only; additional workspaces need a subscription before you can work in them.

Owners can rename or delete a workspace from its settings.

To add teammates to your workspace:

1. Open Members from the account menu.
2. Press the invite button, enter one or more email addresses (comma-separated), and pick a role.
3. Invitees get an email link to join.

Pending invites can be resent or revoked from the same page.

Every workspace member has one of three roles.

• Owner: everything, including billing, workspace settings, and members, plus all Editor abilities.
• Editor: create and run checks, use Compose, and manage the style guide and content.
• Viewer: read-only access to results, guidelines, and insights.

You pick a role when you send an invite, and can change it later from the Members page.

Two-factor authentication adds a second step to your login: a 6-digit code from an authenticator app on your phone.

To set it up:

1. Open Security from the account menu and press Add authenticator.
2. Scan the QR code with your authenticator app, or enter the key manually.
3. Enter the 6-digit code to verify.
4. Save your 8 recovery codes somewhere safe.

Each recovery code works once. If you lose your device, a recovery code gets you back in.

Owners can make two-factor authentication mandatory for everyone in the workspace.

Toggle Require two-factor authentication on the Members page. Once it's on, every member must have 2FA set up before they can access the workspace.

Members who haven't set up 2FA yet are prompted to do so; the article "Two-factor authentication (2FA)" walks through the steps.

Owners can set the workspace's default language on the Members page. Available languages are English, German, French, and Spanish.

The default applies to brands created after the change; existing brands keep their setting. Each brand also has its own locale, so a workspace can hold brands in different languages.

Credits are STETLY's single usage meter. Every AI action (a check, a Compose draft, a style guide build) uses credits from your workspace's monthly allowance.

The meter in the top bar shows your usage at a glance. Billing has the full breakdown, including per-brand usage if your workspace runs several brands.

Your allowance renews with each billing period.

STETLY has three self-serve plans, all at launch prices:

• Starter: €30/month with 850 credits. 1 brand, 2 seats.
• Pro: €90/month with 2,600 credits. 3 brands, 10 seats, includes Brand Insurance.
• Business: €240/month with 7,200 credits. 8 brands, 25 seats, bring your own AI keys.
• Enterprise: custom pricing for teams outgrowing Business. More brands or seats, higher volume, security review, DPA, and annual invoicing.

Prices are flat per workspace, not per seat. Adding teammates doesn't change your bill.

Annual billing gives you 2 months free: pay for 10 months, get 12.

Style guide generation runs are also limited per plan.

As a rough guide: a check costs about 11 credits, and a full style guide build about 310. Actual costs vary with content length.

Rechecks, rewrites, Compose actions, and Stet messages cost less than a full check.

The Billing page shows exactly where your credits went this period, so you never have to guess.

Workspaces on the Business plan that use their own AI keys (BYOK) aren't metered in credits for those calls.

Your first workspace gets a 7-day free trial with full access. No card required.

The top bar shows how many days remain. If the trial ends before you subscribe, the workspace goes into read-only mode: your data, guidelines, and history are all kept.

Everything unlocks the moment you subscribe.

Subscribe or switch plans on the Billing page. Checkout and payment run through Stripe.

At checkout you pick monthly or annual billing. Annual gives you 2 months free.

Manage subscription opens the Stripe billing portal, where you'll find invoices, payment methods, and cancellation.

Only workspace owners can manage billing.

Auto-recharge keeps a busy month from stalling your team.

Owners can turn on Automatically recharge credits on the Billing page. When your monthly allowance runs out, additional credits are billed at the end of the period.

You set a cap, so recharging never exceeds the limit you're comfortable with.

Business workspaces can plug in their own model provider API keys (Anthropic, OpenAI, and others) on the Billing page.

Calls made with your keys bill your provider account directly. They aren't deducted from your STETLY credits.

Connect Instagram and your knowledge library stays in sync with what you actually publish. Captions, reels, and stories all feed into your brand voice.

You need an Instagram professional account (Business or Creator) linked to a Facebook Page. The connection runs through Facebook Login, which is also what lets STETLY ingest Stories.

1. Open Brand Settings and go to Integrations.
2. Press Connect Instagram.
3. Approve access in the Facebook dialog.

Posts, reels, and stories sync in, captions included. Video content is transcribed, so the way you talk on camera counts toward your brand voice too.

Syncs run automatically on a schedule you can adjust (hours between syncs). You can trigger a manual sync anytime, and Backfill posts pulls in history from before you connected.

Disconnecting stops the sync. Content that already made it into your library stays there until you delete it.

Point STETLY at a Google Drive folder and the documents inside sync into your knowledge library, embedded images included.

1. Open Brand Settings and go to Integrations (you can also start from Add Content).
2. Press Connect Google Drive and sign in.
3. Pick the folder you want to sync.

Re-syncs run on a schedule, so new and updated documents flow in on their own. You can also trigger a sync manually whenever you need the latest version right away.

STETLY can read your Notion pages and keep them in your knowledge library.

1. Create a Notion integration and copy its token.
2. In Brand Settings, open Integrations and connect Notion with that token.
3. In Notion, share the pages you want STETLY to read with your integration.

Only shared pages sync. They stay updated incrementally, so edits in Notion show up in your library without a full re-import.

A personal access token (PAT) lets your own tools talk to the STETLY REST API, for example to submit content for checks straight from your CMS or CI pipeline. The same tokens also power the n8n and Make recipes and the developer path into the MCP connector.

1. Open the avatar menu (top right) and choose API tokens & connected apps, then press New token.
2. Name it and pick its scopes: run checks, rewrite drafts, draft via the MCP tools, or read brand and knowledge data. A token can cover the whole workspace or just a single brand.
3. Set an expiry, up to 365 days out.
4. Copy the token right away. For security it's shown only once.

Send the token as a bearer token on your API requests. If a token leaks or you stop using it, revoke it from the same page and it stops working immediately.

If you missed the one-time reveal, there's no way to see the token again. Just create a new one and revoke the old.

The same page also lists Connected apps: AI assistants like Claude or ChatGPT you've connected over MCP. Disconnecting one there cuts its access within the hour.

STETLY speaks the Model Context Protocol (MCP), so AI assistants and editors can work with your brand directly. Once connected, Claude, ChatGPT, or Cursor can list your brands and content types, draft new content in your voice, revise drafts, check any text against your brand rules, search your past content, and pull your check history and insights for status reports, all from where you're already writing. Results include links back into STETLY, so the full history and review tools are one click away.

To connect claude.ai, ChatGPT, or Cursor:

1. Add the STETLY server URL where your tool expects it: claude.ai and ChatGPT call it a custom connector in their settings, Cursor calls it an MCP server. You can copy the URL from Brand Settings → Automation & AI under Claude, ChatGPT & Cursor (MCP).
2. Your browser opens a STETLY consent page. Sign in if you aren't already.
3. Review what the tool will be able to do, pick one brand or all brands in your workspace, and press Connect.

That's it. The connection shows up under API tokens & connected apps (avatar menu, top right), where you can disconnect it at any time. Disconnecting cuts the tool's access within the hour.

For Claude Code or your own scripts, you can skip the consent flow and use a personal access token as the bearer header instead. See the personal access tokens article for how to create one.

Two things to know:

• Drafting and checking through a connector uses your workspace credits, exactly like Compose and Checks in the app.
• The assistant only reaches the brand you picked at connect time. To change it, disconnect and connect again.

If your content moves through n8n, STETLY can QA it on the way: every draft your workflow produces gets a brand check before it ships.

You need a personal access token with the check:run scope. Create one under API tokens & connected apps (avatar menu, top right) and copy it when it's shown.

Set up n8n once:

1. Create a credential of type Header Auth. Header name Authorization, value Bearer obx_pat_... (your token).
2. Add an HTTP Request node to your workflow. Method POST, and paste your brand's check endpoint as the URL. You can copy the exact endpoint from Brand Settings → Automation & AI under n8n.
3. Send a JSON body with your content, and pick the Header Auth credential you created.

The response carries the findings your workflow can act on: route them to Slack, gate a publish step on the score, or write them back to Notion or Airtable.

A common starter workflow: a schedule trigger lists new drafts from your CMS, the HTTP Request node checks each one, and an IF node posts to your review channel only when findings come back.

Each check uses workspace credits, the same as running it in the app. If the token stops working, check that it hasn't expired or been revoked, then mint a new one.

Make scenarios can send content through a STETLY brand check with a single HTTP module, so anything your scenario produces or picks up gets checked before it goes out.

You need a personal access token with the check:run scope. Create one under API tokens & connected apps (avatar menu, top right) and copy it when it's shown.

In your scenario:

1. Add an HTTP → Make a request module. Method POST, and paste your brand's check endpoint as the URL. You can copy the exact endpoint from Brand Settings → Automation & AI under Make.
2. Add two headers: Authorization with value Bearer obx_pat_..., and Content-Type with value application/json.
3. Set the body type to raw JSON and map your content into it.

Tip: keep the token in a Make data store and reference it in the header, rather than pasting it into the scenario blueprint itself.

The response comes back as one bundle with the check result. Use an Iterator to expand the findings array into one bundle per finding, then route them to email, Slack, or whichever module handles your reviews.

Each check uses workspace credits, the same as running it in the app.

Open Style Guide and add rules for your brand: run the automatic generator or add guidelines by hand. Then come back to Run a Check and it will go through.

The message appears because a check needs something to check against. Without guidelines there's no definition of "on-brand" yet, so STETLY asks you to set one up first.

Open the Errors tab in the findings. It shows the blocking rule that failed the check.

A blocking rule fails a check no matter how high the score is. These are usually contract or legal obligations, the kind of thing where 95 out of 100 isn't good enough. Fix the flagged issue and recheck, and the result will match the score again.

Submit the real draft and you'll get a score. The check you ran was judged not evaluable because the content looked like placeholder text rather than actual content, and the credit for it was refunded.

STETLY makes this call on purpose. Scoring "test test test" against your brand rules would produce a meaningless number, so a placeholder check returns no score and costs you nothing. The article on why a check comes back "not evaluable" has more detail.

You can close the wizard. The run keeps going in the background, and the Dashboard shows a banner where you can pick it back up.

Reading a large corpus simply takes a few minutes. The generator works through everything you've ingested, so more content means a longer run.

If a run fails, open it and press Retry. Failed runs don't count against your monthly runs.

You have three options, all managed by a workspace owner on the Billing page:

• Wait for your monthly reset. The date is shown on the Billing page.
• Enable auto-recharge so extra credits are added automatically when you run low.
• Upgrade to a plan with a bigger monthly allowance.

Nothing else changes while you're out of credits. Your content, guidelines, and history are all still there; only actions that spend credits are paused.

Subscribe from the Billing page and everything unlocks immediately.

The banner means your trial ended without an active subscription. Nothing was deleted: your brands, guidelines, and check history are all intact, just read-only until a plan is active.

Check two things, then reconnect from Brand Settings under Integrations:

• Your Instagram account is a professional account (Business or Creator).
• It's linked to a Facebook Page you manage.

Meta requires both for API access, so a personal Instagram account or an unlinked one will fail every time. You can switch to a professional account and link a Page in the Instagram app's settings, then try the connection again.

Two things to check:

1. Open the Members page and look at their invite status. If it's still pending, resend the invite; the original may have expired or landed in spam.
2. If your workspace requires two-factor authentication, they need to enable 2FA under Security in their own account before they get in.

Once the invite is accepted and 2FA (if required) is on, access is immediate.

Sign in with one of your recovery codes: on the verification screen, choose Use a recovery code instead and enter a code you saved when you set up 2FA.

Once you're in, go to Security, remove the old authenticator, and add a new one on your current device. Your old recovery codes are replaced when you do.

Lost the recovery codes too? Contact support at [email protected] and we'll verify your identity and help you regain access.

Grammar tools check whether your writing is correct. STETLY checks whether it's yours: your voice, your rules, your facts, your contract obligations.

A sentence can be flawless English and still be off-brand, off-message, or in breach of a sponsorship contract. That's the layer STETLY covers.

Most teams use both. Grammarly cleans up the prose, STETLY is the brand gate before publishing.

Yes. Each workspace is fully isolated, and within a workspace each brand's data is separated as well.

Your content is used only to check and generate for your own brand. It is never used to train AI models.

Data is hosted in the EU (Frankfurt). The full details are in our privacy policy at /privacy.

Yes, at whatever level you need:

• Delete individual content items from your knowledge library.
• Delete a brand, which removes its profile, rules, content, and history.
• Delete an entire workspace.

For account-level removal, follow the steps at /data-deletion.

Two ways back in, both from the sign-in screen:

1. Use the password reset link. We email you a link that lets you set a new password.
2. Skip the password entirely: request a magic link and we email you a one-click sign-in link instead.

If you signed up with Google, there's no STETLY password to reset. Use Sign in with Google.

If the email doesn't arrive within a few minutes, check your spam folder and confirm you're using the address you registered with. Still stuck? Email [email protected].