SEObot

Deep inside one class: publishing. Publishing destinations are numerous and each has its own documented integration page: WordPress, Webflow, Shopify, Ghost, HubSpot, Notion, Squarespace, Wix, Unicorn Platform and custom sites via REST API or the Next.js package, plus a Framer marketplace plugin and a Zapier route. Every one of them is a place to publish.

There is no other class, with no analytics connection, no Google Search Console or Analytics integration documented anywhere, no CRM, no keyword data provider and no backlink tool. The analytics gap is worth stating specifically rather than as a generic absence: a product that publishes for search visibility and cannot read Search Console cannot close its own loop.

Sourceseobotai.com/integrations and seobotai.com/integrations/apiread 2026-09-02

A complete unattended pipeline ending in published content, and it is the product's own fixed sequence.

Onboarding is a URL and a button, after which SEObot researches the site, audience and keywords, clusters keywords by search pattern and intent, builds a content plan, writes articles every week with fact checking and citations, links each new article into the existing corpus, and publishes to the customer's CMS.

The buyer can give ideas or take control, but no flow the buyer designs or changes, no branching or conditions the buyer configures, no set of coordinated agents and no runtime or API over the flow is documented; programmatic SEO is a second generation mode, not a flow surface.

Sourceseobotai.com and seobotai.com/seo-ai-agentread 2026-10-01

Real and multi-source, but the sources are the open web rather than the customer's own knowledge. The agent researches the site, audience and keywords at onboarding, clusters keywords by search pattern and user intent rather than working from flat lists, analyses competitor content, reads the existing published corpus in order to link new articles into it, and accepts YouTube video URLs as raw input.

A fact-checking system with citation of sources sits over generation, so claims carry provenance. There is no customer-owned grounding structure, with no knowledge base, no brand voice or style object, no document ingestion, no product documentation input and no retrieval layer over the customer's own material. The vendor's own FAQ states that article quality varies against human writers, sometimes matching, sometimes falling short and occasionally exceeding.

Sourceseobotai.com and seobotai.com FAQread 2026-09-02

An approval step the vendor ships, off by default. The home page says "You can always ask SEObot to email you a list of the articles to accept/decline", that articles can be edited in SEObot before publishing, and that a user can give it ideas or take control, while stating it "runs 100% autopilot by default". Accept or decline before an article is published is a person approving before the action commits.

The default matters to a buyer: without the setting, articles publish to the customer's live site under their brand without review. An anti-hallucination and fact checking system with citations constrains what is written and sits beside the gate. The approval flow is described on the home page rather than documented, and how to switch it on is not published.

Sourceseobotai.comread 2026-10-01

A privacy policy and terms, and nothing else. The policy is not wholly boilerplate, carrying a product specific clause on domain associated data covering generated articles, metadata, keyword data, analytics and publishing history, how it moves when a domain transfers between accounts, and that the prior holder's access terminates.

Beyond that there is no certification of any kind, no trust center, no security page, no DPA, no subprocessor list, no encryption statement, no SSO, no role based access and no penetration test evidence. The policy does establish two things. The vendor declares itself a data controller of user data rather than a processor acting on instruction, which is the opposite of the arrangement enterprise buyers expect.

And it discloses that domain and content data is processed using third party AI services, without naming them. This is a self serve product from $49 a month built by an independent developer, so an enterprise security program would be surprising to find.

What the absence still means is real: the tool holds publishing credentials to the customer's live public website and posts under their brand autonomously by default, which is a standing access grant with no documented scoping, controls or revocation process around it.

Sourceseobotai.com/privacy-policy and seobotai.com/terms-of-serviceread 2026-09-02

Nothing SEObot itself exposes records what the agent did: no audit log, no run history, no decision trail explaining why a topic was chosen or which keyword cluster drove it, no record distinguishing agent-written text from human edits, and no retention policy.

The customer's CMS records what was posted and when, and outbound webhooks push events for the customer to store in systems they build and operate, but those are the customer's own records rather than the vendor's. The opt in email listing articles for approval is a notification rather than a log.

That gap is sharper here because the default is autonomous publication to a live public site under the customer's brand, so the question of why a given article exists is one a buyer may need answered after the fact and cannot be.

Sourceseobotai.com/integrations/webhooks and seobotai.com FAQread 2026-09-02

Accumulated output, no agent memory.

Each new article is linked into the growing corpus of everything already published, so the system tracks what it has produced, what each piece targets and how the internal link graph is shaped, and the privacy policy describes domain-associated data covering generated articles, metadata, keyword data, analytics and publishing history.

That is the product's record of its own output rather than agent memory: no memory construct, no facts about the user or brand carried across sessions as agent state, and no retention or forgetting control at the agent level is documented.

Sourceseobotai.com and seobotai.com/privacy-policyread 2026-10-01

A stated jurisdiction and nothing the customer selects. The privacy policy states that if a user is located outside the United States, the vendor transfers their data, including personal data, to the United States and processes it there. No region selection, EU option, private deployment, VPC or self hosted route is offered.

Sourceseobotai.com/privacy-policyread 2026-10-01

More than one agent ships, but there is no catalog. The core SEO robot is joined by a separately named Listing Bot automating product listings, the vendor's own framing is plural at SEO AI agents rather than one assistant, programmatic SEO is offered as a distinct mode building templates at scale, and a Framer marketplace plugin packages the capability for that platform's users.

A set of free public tools sits alongside, covering FAQ, Q and A and article schema generation, canonical checking, robots.txt generation, broken link checking, meta description and title generation and sitemap extraction, which is a published toolset a user picks from even though it sits outside the agent. There is no catalog of agents, no template library, no blueprints a customer browses and instantiates, no marketplace and no way to compose, clone or share an agent. Article structures are generated from keyword and intent clustering rather than selected from prebuilt templates.

Sourceseobotai.com/seo-ai-agent, seobotai.com/tools and the Framer plugin listingread 2026-09-02

A recurring cadence wakes the agent with no one asking. The home page says SEObot "will make a content plan and start producing articles every week" and that it "runs 100% autopilot by default", so work reaches the agent on its own schedule after onboarding.

Coverage is one destination class, the customer's website through nine CMS integrations (Framer, Ghost, HubSpot, Notion, Shopify, Unicorn Platform, Webflow, Wix, WordPress) plus REST API, webhooks and a Next.js package, in more than 50 languages; outbound webhooks notify rather than start work. The cadence is described rather than configured on a documented surface, and whether the schedule can be changed is not stated.

Sourceseobotai.com and seobotai.com/integrationsread 2026-10-01

No provider named and no customer choice, though the vendor acknowledges third parties are involved. The privacy policy states that the vendor processes domain and content data using artificial intelligence and third-party AI services for the purpose of generating content, which confirms the dependency exists and is external while naming nobody. Nothing offers model selection, routing, bring your own model or bring your own key, and pricing is by article volume and plan tier rather than by tokens or credits, so a buyer sees no signal about model cost or identity at any point.

Sourceseobotai.com/privacy-policy and seobotai.com FAQread 2026-09-02

A published SDK with public documentation. The vendor's API integration page documents where a customer finds their API key, points to a published npm package named seobot, links documentation in a public GitHub repository under MarsX-dev, and provides a working Next.js integration example in a second repository. It also publishes a demo API key so a developer can build and test an integration before buying anything.

Alongside the SDK sit a documented webhooks page delivering real-time events, a Framer marketplace plugin, and dedicated integration pages for WordPress, Webflow, Shopify, Ghost, HubSpot, Notion, Zapier, Squarespace and Unicorn Platform. The npm package is the SDK, the GitHub repository is the reference, and the API key location is documented. There is no MCP server, no developer portal beyond the repositories, and no partner or certification program.

Sourceseobotai.com/integrations/api, the seobot npm package and github.com/MarsX-dev/seobotread 2026-09-02

A checkable artifact, and nothing that tests the agent.

Inside the pipeline a fact checking and anti-hallucination system cites its sources, so a reviewer can follow a citation rather than trust a claim, and the vendor's own FAQ says article quality varies against human writers and sometimes falls short, which is the vendor's statement of where review matters. That checks the customer's artifact.

No sandbox or dry run, no scoring of output quality, no A/B testing of article variants, no regression check when configuration changes, and no measurement of ranking or traffic outcomes inside the product.

Sourceseobotai.com and seobotai.com FAQread 2026-10-01

SEObot publishes through documented CMS integrations and APIs into WordPress, Webflow, Shopify, Ghost, HubSpot, Notion, Squarespace, Wix and custom sites via its REST API and npm package, all of which expose programmatic publishing routes, so no interface-level operation is required and none is documented.

Its research inputs are keyword and search data plus crawls of the customer's own site and of competitor pages, which is content ingestion from public web pages rather than operating software on a user's behalf. No browser control, headless session, screen operation or robotic process automation appears anywhere on the property.

Sourceseobotai.com/integrations/api and seobotai.comread 2026-09-02