Instructions in this file are plain Markdown (no hidden or encoded content).

Bundle version: 1.3.1

SEO Expert: User Guide#

What this skill is#

This skill equips your OpenClaw agent with structured guidance for search engine optimization: technical foundations, on-page content, information architecture, performance signals relevant to crawl and indexation, and measurement (without guaranteeing rankings).

References: German Google Search Central excerpts under {baseDir}/references/ — fundamentals ( 01_seo_grundlagen.md), crawling & indexing ( 02_crawling_indexierung.md), ranking & appearance ( 03_ranking_darstellung.md hub + 03.1– 03.7), monitoring & troubleshooting ( 04_monitoring_fehlerbehebung.md), web-specific guides ( 05_webspezifische_leitfaeden.md hub + 05.1– 05.8). Index: {baseDir}/references/OVERVIEW.md.

Not exhaustive: SEO platforms, algorithms, and guidelines change continuously. Prefer primary sources (e.g. Google Search Central, Bing Webmaster, your stack’s documentation) when behavior, limits, or compliance matter.

The AI may use curl for HTTP checks (status, headers, robots) when your gateway policy allows. Broader tools follow your host configuration; optional OpenClaw policy notes can be added under {baseDir}/references/ as the skill grows.

No companion plugin (v1.1)#

There is no bundled OpenClaw plugin in v1 — only this skill and tools your host already allows.

Installation (typical)#

1. Install the skill (ClawHub or skills/seo-expert-pro on the gateway host).
2. Allow curl (and optionally browser or fetch tools) per your openclaw.json / sandbox policy.
3. Restart the gateway after skill or policy changes if needed.

Details: {baseDir}/README.md, {baseDir}/references/OVERVIEW.md.

What you can expect from the AI#

• Structured audits and checklists grounded in documented SEO practice — not black-hat or platform-terms-violating tactics.
• No secrets in chat — API keys for Search Console or third-party tools belong in env only (see {baseDir}/references/AUTH.md when present).
• Honest limits: nobody can ethically promise rankings; the skill focuses on discoverability, quality, and measurement hygiene.

Required setup (eligibility)#

Per metadata.openclaw.requires:

1. curl available on the gateway for optional HTTP diagnostics.

Optional: target site URL or API credentials — document env names in .env.example and future AUTH.md when you add tool integrations.

When the agent should use this skill#

Use for SEO-related work: crawling/indexing, metadata, structured data, site structure, redirects, Core Web Vitals context, analytics/search-console style workflows, content quality for search — and for clarifying what is vs is not in scope.

Load {baseDir}/references/OVERVIEW.md first, then the smallest matching reference ( 01_…, 02_…, 04_…, 03 hub / 03.n_… parts, or 05 hub / 05.n_… parts).

Rules for the assistant (summary)#

1. Cite primary guidelines where possible; flag uncertainty when platforms differ or update silently.
2. Never promise rankings, “instant” results, or tactics that violate search engine or legal terms.
3. Never echo tokens or API keys; reference env names only.
4. Prefer measurable recommendations (what to check, how to verify) over vague advice.

Where work runs: on the OpenClaw gateway and in conversation — not inside Google/Bing systems unless the user connects official tools separately.