01
Overview
SEOAgent v1.0 is a codebase-first SEO harness designed to work with coding agents rather than replace them. Instead of living only in a browser dashboard, it installs a Skill and command-line tool in a website repository so tools such as Claude Code, Cursor, Codex and Grok bots can audit and edit the actual site files.
02
What is SEOAgent?
SEOAgent describes itself as an SEO engine for coding agents. Its local layer gives an agent structured commands and project context for crawling a site, checking technical SEO, planning pages, improving metadata and schema, finding internal-link opportunities, and preparing SEO content as ordinary repository changes.
The important architectural point is that SEOAgent is not the language model. The user's coding agent does the reasoning and file editing; SEOAgent supplies the SEO workflow, evidence and synchronization layer.
03
What launched in SEOAgent v1.0?
The September 16, 2026 v1.0 release adds first-class Grok bot support alongside Claude Code, Cursor and Codex, Open Knowledge Format publishing for AI-search readiness, and product screenshot capture from a customer's codebase. The public site and developer docs now identify the product as v1.0.
04
How the local CLI works
The documented install command is npm install -g @seoagent-official/seoagent, followed by seoagent init inside a repository. An npx workflow is also documented for running without a global install.
The CLI creates a .seoagent/ workspace. Commands include status, doctor, crawl, keywords, internal-links, okf, llms, sitemap, and cloud-connected commands such as login, sync and inbox.
SEOAgent says it does not push to the repository or publish to a CMS automatically. Local changes are written by the coding agent and remain reviewable before deployment.
05
Supported coding agents
At launch the product lists Claude Code, Cursor, Codex and Grok bots as supported coding-agent environments, plus a standalone CLI. Integration depth can differ by environment, so support should not be interpreted as every agent exposing identical capabilities.
06
SEOAgent pricing
The current public pricing page lists the local Skill at $0 and describes it as free forever. The cloud Autopilot plan is listed at $49 per site per month with a seven-day trial.
The free layer runs using the model or coding-agent subscription the user already has. The paid cloud layer adds Google Search Console analysis, competitor and SERP research, a review workflow, sitemap regeneration/submission when GSC is connected, backlink prospecting/draft outreach, and emerging-query/rank monitoring.
Pricing is time-sensitive and should be rechecked before publication or purchase.
07
Google Search Console and the feedback loop
The cloud product connects to Google Search Console and uses performance data to propose updates. This is more interesting than a one-shot AI writer because the stated workflow is measure → suggest → approve → sync → implement, allowing actual search evidence to influence later changes.
That does not mean the tool guarantees ranking improvements. Search performance depends on the site, market, content quality, links, technical state and search-engine behavior.
08
Open Knowledge Format and llms.txt
SEOAgent can generate llms.txt and an Open Knowledge Format bundle. Its documentation describes OKF as a directory of Markdown files with YAML frontmatter based on an open specification published by Google Cloud. The goal is to expose structured organizational knowledge in a form AI systems can consume.
SEOAgent writes these files into the repository rather than silently deploying them. Existing route ownership is checked to reduce the risk of overwriting a separately managed llms.txt implementation.
Publishing OKF or llms.txt should not be described as a guarantee of citations, rankings or inclusion in AI answers. They are machine-readable discovery/context mechanisms, not a contractual distribution channel.
09
Product screenshot capture
Version 1.0 also adds screenshot capture from the customer's own application codebase. SEOAgent identifies where a product image would help, renders the relevant interface, stores the resulting image in the project and adds descriptive alt text. If capture is not possible, the vendor says it reports the gap rather than substituting a generated placeholder.
10
Human approval and automation
The standard local workflow is explicitly reviewable: the coding agent edits files, the user reviews the diff and then decides whether to commit and deploy.
Grok-bot documentation describes an optional more autonomous cloud queue with auto-approve enabled by default in that specific workflow. Teams should review this setting before connecting a production repository rather than assuming all SEOAgent modes have the same approval behavior.
11
Who is SEOAgent for?
The clearest fit is a founder, SaaS team, developer or SEO operator whose website lives in Git and who already works with a coding agent. It is less naturally suited to teams that expect a traditional no-code CMS-only SEO dashboard and do not want an agent editing repository files.
12
Limitations and verification points
Independent adoption and outcome evidence is still limited because v1.0 is new. SEOAgent states that its SOC 2 program is in progress; that is not the same as completed SOC 2 attestation. Claims about improved AI visibility from OKF or llms.txt should be treated as intended benefits, not guaranteed outcomes.
Before production use, verify the current npm package, repository permissions, cloud data handling, approval settings, GSC scope and the exact files the Skill proposes to modify.
13
FAQ
Is SEOAgent free?
The current site lists the local Skill as free. Its Autopilot cloud plan is listed at $49 per site per month.
Does SEOAgent publish changes automatically?
The standard local workflow does not. It creates reviewable repository changes. Some autonomous bot workflows can use pre-approved queues, so configuration matters.
Does SEOAgent work with Codex?
Yes. The current documentation lists Codex alongside Claude Code, Cursor and Grok bots.
What is SEOAgent OKF?
It is SEOAgent's workflow for generating and validating an Open Knowledge Format bundle: structured Markdown/YAML knowledge intended for machine consumption.
Does OKF guarantee ChatGPT or Google citations?
No. There is no verified guarantee that publishing an OKF bundle or llms.txt will produce rankings or citations.
Does SEOAgent require a separate AI model subscription?
The local Skill is designed to run through the coding agent/model environment the user already has. Cloud features are separately priced.
Sources
Primary and supporting sources
Use the linked documentation to check current pricing, access and product behaviour. The dates above describe this article; they do not guarantee that every vendor claim remains current.