How do I install the Geo Site Diagnose skill?

Run npx skillmds@latest add yaojingang/geo-site-diagnose in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Geo Site Diagnose safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Geo Site Diagnose?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Geo Site Diagnose free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Geo Site Diagnose?

yaojingang (@yaojingang) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Geo Site Diagnose by yaojingang — Agent Skill

Name: Geo Site Diagnose
Author: yaojingang

Geo Site Diagnose

Crawl one public website from a supplied URL, select up to ten representative page types, score evidence-lined GEO readiness, and generate a standalone visual HTML report. Use for 网站 GEO 诊断、全站 GEO 体检、网站 GEO 审计, or requests such as “帮我诊断网站GEO并附上网址”. Excludes single-page or brand-only diagnosis, live AI ranking or citation measurement, authenticated sites, and website modification.

yaojingang Updated Sep 17, 2026

File contents

GEOHub Website Diagnosis

Workflow

Read references/site-diagnosis-method.md, references/crawler-policy.md, and references/scoring-policy.md.
Extract the single public URL. Ask only for the URL when it is missing.
Run python3 scripts/run_site_diagnose.py --url <URL> --output <runs-root> --locale <locale> --max-pages 10 --render auto.
Inspect quality-report.json, diagnostic confidence, source gaps, and limitations before presenting scores.
Validate the report against references/report-module-contract.md and references/report-design-system.md.
Deliver the complete Artifact Bus run directory and open report.html for the user.

Output contract

Return a bounded run containing source snapshots, crawl and sampling manifests, structured site and page diagnoses, evidence ledger, remediation backlog, quality and lineage artifacts, and one offline report.html. Read references/report-contract.md for the file contract and report format boundary.

The offline report runtime is packaged under assets/. Governed trigger and output cases are stored under evals/.

Boundaries

Fetch only public HTTP(S) resources on the approved canonical host. Discovery is bounded to 500 inventory URLs, five sitemaps, ten representative pages, 15 MB, and 90 seconds. Preserve blocked or unavailable pages as source_gap. Scores describe observed readiness signals and never claim live AI ranking, recall, citations, traffic, or business effects.

Do not use this Skill for single-page diagnosis, brand-only diagnosis, authenticated sources, live citation measurement, or website modification.

---name: geo-site-diagnosedescription: Crawl one public website from a supplied URL, select up to ten representative page types, score evidence-lined GEO readiness, and generate a standalone visual HTML report. Use for 网站 GEO 诊断、全站 GEO 体检、网站 GEO 审计, or requests such as “帮我诊断网站GEO并附上网址”. Excludes single-page or brand-only diagnosis, live AI ranking or citation measurement, authenticated sites, and website modification.---# GEOHub Website Diagnosis## Workflow1. Read `references/site-diagnosis-method.md`, `references/crawler-policy.md`, and `references/scoring-policy.md`.2. Extract the single public URL. Ask only for the URL when it is missing.3. Run `python3 scripts/run_site_diagnose.py --url <URL> --output <runs-root> --locale <locale> --max-pages 10 --render auto`.4. Inspect `quality-report.json`, diagnostic confidence, source gaps, and limitations before presenting scores.5. Validate the report against `references/report-module-contract.md` and `references/report-design-system.md`.6. Deliver the complete Artifact Bus run directory and open `report.html` for the user.## Output contractReturn a bounded run containing source snapshots, crawl and sampling manifests, structured site and page diagnoses, evidence ledger, remediation backlog, quality and lineage artifacts, and one offline `report.html`. Read `references/report-contract.md` for the file contract and report format boundary.The offline report runtime is packaged under `assets/`. Governed trigger and output cases are stored under `evals/`.## BoundariesFetch only public HTTP(S) resources on the approved canonical host. Discovery is bounded to 500 inventory URLs, five sitemaps, ten representative pages, 15 MB, and 90 seconds. Preserve blocked or unavailable pages as `source_gap`. Scores describe observed readiness signals and never claim live AI ranking, recall, citations, traffic, or business effects.Do not use this Skill for single-page diagnosis, brand-only diagnosis, authenticated sources, live citation measurement, or website modification.

Frequently asked questions