Any public site. No onboarding gate.
Give it a domain or URL. Public audits need no client registration, ownership verification or paid data subscription. Without modification access, the work stays at investigation and concrete recommendations.
ClawSEO · Advanced Agentic Discoverability Capability
Direct ClawSEO through a connected channel. Measure public websites, retain project context, delegate approved work to configured Codex tools and verify outcomes.
Ask why visibility dropped, where a service business is losing discovery opportunities, or how a product can be easier to find and understand. ClawSEO investigates the evidence available and develops the next useful action.
Give it a domain or URL. Public audits need no client registration, ownership verification or paid data subscription. Without modification access, the work stays at investigation and concrete recommendations.
ClawSEO chooses available tools, establishes a baseline, follows hypotheses and prioritizes defensible improvements. You provide the objective rather than operating its tools.
Once the evidence identifies a useful change, connect the relevant hosting, CMS, repository or business system. ClawSEO can hand a scoped assignment to configured Codex execution, track its status and inspect the result. You control the permitted changes.
Check the live result and repeat the same benchmarks. Track findings, experiments and unresolved issues. Stop when further changes lack evidence or would harm the property.
Use your preferred connected channel to set objectives, follow assignments and approve changes. Telegram and WhatsApp are supported entry points, with WhatsApp requiring phone pairing. Other OpenClaw channels depend on compatible setup and permissions.
Run local Lighthouse to establish performance, accessibility and SEO baselines, then measure again after a relevant change. Reuse recent results where appropriate and respect provider quotas. Background checks focus on meaningful developments rather than repeatedly consuming benchmarks.
Search, local services, ecommerce and AI discoverability share the same evidence-led workflow. Scores never outrank usability, design, accessibility, factual accuracy, trust or business performance. Technical improvements alone do not prove ranking or revenue gains.
Public research can begin with any public domain. Saved objectives, findings, decisions and linked assignments keep the operation in context across conversations. Connected analytics and business accounts add private evidence when access is available.
The current complete experience is verified on Mac mini. ClawSEO brings the work together through a local operator and a natural conversation.
ClawSEO turns an objective into an evidence-led SEO operation. Audit any public domain, investigate useful findings, develop a solution and implement it when the relevant systems are connected. The current full operator workflow is verified on macOS.
curl -fsSL -A "Mozilla/5.0 ClawSEO-Installer/0.1" https://clawseo.com/install.sh | bashThe installer prints absolute paths you can copy immediately. Use install-daemon on macOS or native Linux. In WSL, use no-install-daemon and activate with no-service. For public auditing, provide the domain you want to investigate. Domain ownership is not a requirement for public reads. Activation configures a default working target rather than granting private account or modification access.
"$HOME/.clawseo/bin/clawseo-openclaw" onboard --install-daemon --skip-skills --skip-bootstrap"$HOME/.clawseo/bin/clawseo-openclaw" onboard --no-install-daemon --skip-health --skip-skills --skip-bootstrap"$HOME/.clawseo/bin/clawseo" activate example.com "primary service""$HOME/.clawseo/bin/clawseo" activate example.com "primary service" --no-serviceexport PATH="$HOME/.local/bin:$PATH"Supported terminals are macOS and Linux. Windows requires a WSL Linux distribution; native PowerShell and Command Prompt are not supported. WSL uses manual-run mode by default; always-on startup requires a separately configured and verified Windows Scheduled Task.
Organizes page facts through heading paths, fields, dates, units, identifiers, tables, metadata, and sources.
Evaluates whether important information is easy to find and understand.
Analyzes imported repeated AI observations and traces claim origins and source families.
Turns findings into reports, priorities, private drafts, approvals, and verified work.
ClawSEO is not installed directly into ChatGPT or OpenAI. If you choose OpenAI as the model provider, the owner authenticates a supported OpenAI model through the dedicated ClawSEO OpenClaw profile. ClawSEO remains the installed SEO operator on its own independent Mac, PC, or VPS.
No business-system credentials are required to inspect public pages, robots.txt, sitemaps, metadata, canonicals, schema, and AI-crawler signals, then create local reports, briefs, drafts, and an approval queue.
CMS changes and publishing, Search Console or analytics access, email or outreach, business-profile edits, review replies, and other external operations require your access and approval.
An external action runs only through a configured executor after the exact item is approved, then an independent verifier checks the real result before completion.
ClawSEO puts evidence location, retrieval testing, AI visibility distributions, provenance, and traditional technical and content SEO into one operating loop. It does more than describe a gap. It organizes findings into prioritized work while preserving the evidence and result checks.
The complete operator runs on a separate Mac, PC, or VPS of its own. The methodology comes from @organicrankings and more than 21 years of personal SEO experience. iLuvSEO supplies written reference knowledge from the same author.
Puts the ClawSEO SEO operator on its own independent Mac, PC, or VPS.
Gives ClawSEO its own independent Mac, PC, or VPS, SEO workspace, state, approvals, and operating loop.
Gives the organization a dedicated SEO operating loop instead of scattered prompts, agency handoffs, or one-time audits.
The owner authenticates a supported provider through the dedicated ClawSEO OpenClaw profile while the local operator owns the workflow.
Checks crawl access, robots.txt, sitemap.xml, llms.txt, titles, descriptions, H1s, canonicals, and schema.
Records technical fixes, on-page work, content briefs, authority opportunities, brand tasks, and reputation drafts.
Runs an approved external action only through an owner-authorized execution adapter.
Checks the real result before marking an action complete.
Readiness checks confirm the installation can observe, act, verify, and produce useful SEO output.
Keeps queued work, reports, and approvals organized for the owner or team.
Blocks public actions until the user approves the exact item.
Turn each SEO expert operator on or off per install; disabled operators are skipped by the agent cycle.
Uses launchd on macOS and a systemd user timer on Linux, with an explicit manual-run path when neither is available.
Evaluates whether a business and its information can be found, understood and verified by search systems and AI agents.
Checks response health, crawlability, robots.txt, sitemap.xml, llms.txt, canonicals, schema, and page signals.
Expands seed terms, groups intent, and creates clear content-brief actions.
Searches source-linked iLuvSEO textual knowledge and applies the @organicrankings method with current-fact validation.
Audits owned site assets, sitemap pages, metadata, H1, canonical, and schema coverage.
Turns audits and keywords into title, description, heading, internal-link, and schema tasks.
Builds legitimate citation, partner, resource, expert-source, and brand-mention queues with approval gates.
Prepares entity-consistency tasks, real-review response drafts, profile-correction queues, and reputation-risk reports.
Complete installation and activation first, then use the command that matches the job. This is a reference for operators, not a substitute for the product flow above.
curl -fsSL -A "Mozilla/5.0 ClawSEO-Installer/0.1" https://clawseo.com/install.sh | bash -s -- install-appliance example.com"$HOME/.clawseo/bin/clawseo-openclaw" onboard --install-daemon --skip-skills --skip-bootstrap"$HOME/.clawseo/bin/clawseo-openclaw" onboard --no-install-daemon --skip-health --skip-skills --skip-bootstrap"$HOME/.clawseo/bin/clawseo" activate example.com "primary service"export PATH="$HOME/.local/bin:$PATH"clawseo activate example.com "primary service" --no-serviceclawseo agent-connect-openclaw --command="$HOME/.clawseo/bin/clawseo-openclaw" --agent=mainclawseo agent-circuit-testclawseo agent-readiness --require-proofclawseo agent-operate "seo agency" --execute-drafts --limit=3clawseo agent-run --onceclawseo agent-operatorsclawseo agent-operator-disable offpage-authorityclawseo agent-operator-enable offpage-authorityclawseo agent-operator-toggle brand-reputationclawseo agent-keyword-research "seo agency"clawseo agent-property-audit example.comclawseo agent-evidence-audit example.com --limit=25clawseo agent-evidence-query "exact fact" --index=/path/to/evidence-audit.json --method=structured-hybridclawseo agent-visibility-analyze --exampleclawseo agent-visibility-analyze /path/to/visibility-observations.jsonclawseo agent-provenance-audit --exampleclawseo agent-provenance-audit /path/to/provenance-claims.jsonclawseo agent-operator offpage-authorityclawseo agent-statusclawseo agent-queueclawseo agent-approve <id>clawseo agent-execute <id>clawseo agent-execute --approvedclawseo agent-retry <id>clawseo agent-reject <id>clawseo agent-reportclawseo agent-install-serviceclawseo agent-uninstall-serviceclawseo doctorclawseo updateclawseo uninstallclawseo uninstall --purgeclawseo wheredoctor verifies package and endpoint integrity only. Run agent-readiness --require-proof to check the model, owned property, executor, verifier, and useful output. Normal uninstall preserves reports, queues, deliverables, and workspace owner context; --purge also removes the dedicated ClawSEO profile and all retained data.
Need installation handled by us?See plans
Install the local operator, connect a model and give ClawSEO a website and objective.