Cogbird is now in private beta.

Terms of Service

What you agree to by using the service, what we promise, and what we do not.

Last updated

1. Agreement

Cogbird is operated by Everace AB (org. nr 559582-9663, VAT SE559582966301), Marieviksgatan 41, 117 60 Stockholm, Sweden (“we”, “us”). By creating an account or using the service you agree to these terms. If you are agreeing on behalf of a company, you confirm you are authorised to do so.

2. Access is by invitation

Cogbird is in early access. Accounts are created only from an invitation we issue, and we may decline a request for access without giving a reason. An invitation is issued to one email address and is not transferable.

3. What the service does

Cogbird reads search performance data you authorise us to read, crawls the public pages of a website you have proven you control, and produces a small number of prioritised recommendations with the evidence behind them. Where a change is shipped and reported, it measures the outcome against a control group of your own pages.

4. What the service does not do

We do not modify your website. Cogbird makes no changes to your live pages, your CMS or your DNS, and nothing is ever written back through the Search Console grant, which is read-only. Every recommendation is executed by you or by software you control, and the decision to act on one is yours alone. We are not responsible for the consequences of a change you choose to make, whether or not we recommended it.

One exception, and only after you switch it on. If you install our GitHub app and switch a site on, we may commit to a branch of that repository and open a pull request against it — by applying a change worked out in advance, or through a coding agent that reads and edits the repository for the changes no fixed rule expresses. The pull request is a proposal you review, and closing it is the whole of the rollback. Every site stays in manual mode, where we write nothing at all, until you change it yourself.

A second switch, separate from the first, lets some of them merge. Turning it on lets us merge, without your review, the changes we write with a fixed rule rather than a coding agent — a page title, a meta description, an internal link, a link repointed at the page it already redirects to, and one sentence added to a section your page already has — and only once your own repository reports a passing check on the pull request. Nothing our coding agent wrote is ever merged this way, at any setting, and a share of the sentences always comes to you. It is off unless you turn it on, it cannot be reached without first switching pull requests on, and turning it off returns every change to your review. We do not deploy your site in any mode; whether a merged commit reaches production is governed by your repository, not by us.

5. Crawling, and the domains we will touch

We fetch pages from a domain only after ownership has been proven — by a DNS record, a file, a meta tag, or a Search Console grant. You must not add a domain you do not control. We crawl politely, respect robots.txt, and cap the number of pages fetched per scan.

You are responsible for the lawfulness of what we are pointed at. If you add a domain you do not control, or one whose content you have no right to have processed, we may suspend the account without notice.

6. Agent access

The service can be reached over MCP with a bearer token scoped to a single site. That token is a credential: anyone holding it can read everything the service knows about that site. Keeping it secret is your responsibility, and you can revoke one at any time. Activity performed with a valid token is treated as yours.

7. Your data and your connections

You keep ownership of everything you provide and everything we derive from it. You grant us the licence needed to operate the service — to store, process and analyse that data in order to produce recommendations for you.

Credentials you connect are stored encrypted and used only to read data for the site they belong to. You can disconnect at any time, which deletes the stored credential.

We may use aggregate, anonymised outcomes — the fact that a class of change succeeded or failed on a site of a given size and shape — to improve the service for all customers. This never includes your domain, your URLs, your queries or your content, and it cannot be traced back to you. See the privacy policy.

8. Payment

Subscriptions are billed per verified site, in advance, through Stripe. A trial converts to a paid subscription automatically at the end of the trial period unless cancelled before it ends. Prices are exclusive of VAT, which is added where applicable.

Cancelling stops the next renewal and leaves the subscription active until the end of the paid period. We do not refund partial periods. If payment fails, access is suspended after a retry period and the account becomes read-only before deletion.

9. What a verdict means, and what it does not

The service measures a shipped change against a control group of your own untouched pages and reports the difference. That is a statistical estimate over a defined window, not a guarantee of causation. Search results are influenced by things nobody in this relationship controls — competitors, seasonality, and the search engines’ own ranking changes.

We make no promise of improved rankings, traffic, or revenue, and no promise that any particular recommendation will help. Where the service cannot measure something, it says so rather than estimating.

10. Acceptable use

You must not:

  • Add a domain you do not control.
  • Use the service to crawl, scrape or extract content you have no right to.
  • Resell or redistribute the service’s output as your own product.
  • Attempt to exceed, evade or disable the rate limits and spend caps.
  • Interfere with the service or attempt to access another customer’s data.

11. Warranties and liability

The service is provided “as is”. We do not warrant that it will be uninterrupted or error-free, or that any recommendation will produce a particular result.

To the extent permitted by law, our total liability for any claim is limited to the fees you paid in the twelve months before the claim arose. We are not liable for indirect or consequential loss, including lost revenue, lost rankings or lost data.

Nothing here limits liability that cannot be limited under Swedish law.

12. Suspension and termination

You may stop at any time by deleting your account. We may suspend or terminate an account that breaches these terms, that endangers the service or another customer, or that we are legally required to close.

Deleting a site, workspace or account starts a 30-day window during which it can be restored. After that it is destroyed permanently, including the record of what was shipped and what it achieved, which cannot be reconstructed.

13. Changes

We may update these terms. Material changes will be notified by email to the address on the account at least 14 days before they take effect. Continuing to use the service after that means you accept them.

14. Governing law

These terms are governed by Swedish law. Disputes are subject to the exclusive jurisdiction of the Swedish courts, with Stockholms tingsrätt as the court of first instance.

15. Contact

Everace AB
Org. nr 559582-9663 · VAT SE559582966301
Marieviksgatan 41, 117 60 Stockholm, Sweden
[email protected]