Stobo Mcp — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Stobo Mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
<!-- mcp-name: io.github.stobo-app/seo-audit --> <!-- docs: https://docs.trystobo.com -->
⚠️ Retired (May 2026). The Stobo SEO/AEO audit service has been shut down and its API is no longer available. This MCP server no longer performs audits — it exposes a single stobo_status tool that reports the retirement and nothing else. Published versions remain installable for archival purposes only. Thank you to everyone who used it.Audit any site for SEO, AEO, and E-E-A-T from Claude Desktop. Get a structured fix brief your developer can drop straight into Claude Code, Cursor, or VS Code and start shipping.
You: stobo this site: example.com
Claude: I just received a complete audit for example.com. SEO, AEO, EEAT,
per-page breakdowns, 30+ checks with scores and details.
Want me to generate a fix brief you can download and drop into
Claude Code, Cursor, or hand off to your developer?Fetching every page manually costs ~386,000 tokens across 16+ calls. It still misses Core Web Vitals, TTFB, and Flesch-Kincaid. Those metrics require server-side computation. Stobo runs everything in one call.
| Manual (web_fetch) | Stobo MCP | |
|---|---|---|
audit_site | 71,000 tk · 6 fetches | 20,000 tk · 1 call |
audit_freshness | 60,000 tk · 20 fetches | 3,000 tk · 1 call |
generate_llms_txt | 35,000 tk · 6 fetches | 3,000 tk · 1 call |
extract_tone | 35,000 tk · 10 fetches | 5,000 tk · 1 call |
| Full suite | 264,500 tk · 56 fetches | 47,000 tk · 9 calls |
82% fewer tokens. And you get metrics Claude cannot compute from raw HTML alone.
Claude receives 30+ scored checks across SEO, AEO, and E-E-A-T. It then offers a fix brief: structured markdown with every failing check, prioritized by impact, with specific fix instructions. Paste it into Claude Code, Cursor, or VS Code and start shipping.
Claude also offers generators for quick wins. Failing robots.txt? Claude generates one. Same for llms.txt, sitemap, and freshness code. The audit-to-fix loop runs in a single conversation.
| Say to Claude | What it does | Tool |
|---|---|---|
| "Stobo this site: example.com" | 30 SEO + 7 AEO checks + E-E-A-T + blog detection across your whole site | audit_site |
| "Audit this article: example.com/blog/post" | 7 SEO + 14 AEO checks on a single article | audit_article |
| "Generate a fix brief" | Structured markdown fix brief from a completed audit. Numbered tasks, priority order, fix instructions. IDE-ready. | generate_fix_brief |
| "Check freshness across example.com/sitemap.xml" | Scan your sitemap for datePublished and dateModified coverage | audit_freshness |
| "Generate a robots.txt for example.com" | AI-crawler-friendly robots.txt. All 21 major bots covered. | generate_robots_txt |
| "Generate a sitemap for example.com" | BFS-crawl up to 200 URLs and return a sitemap.xml | generate_sitemap |
| "Generate freshness code for example.com/blog/post" | JSON-LD snippet with datePublished and dateModified | generate_freshness_code |
| "Check connection" | Diagnostic ping. Also tells you if a newer MCP version is available. | check_connection |
| Say to Claude | What it does | Tool |
|---|---|---|
| "Generate llms.txt for example.com" | Crawl your site and return a spec-compliant llms.txt | generate_llms_txt |
| "Extract brand voice from example.com/blog" | Analyze up to 10 articles and create a persistent voice profile | extract_tone |
| "Rewrite this article: example.com/blog/post" | Audit + tone match + full rewrite in one pipeline | rewrite_article |
| "Audit the UX of example.com" | 50+ UX checks across accessibility, forms, typography, and navigation | audit_ux |
| "How many credits do I have?" | Credit balance and usage breakdown | get_credits |
audit_siteaudit_articlegenerate_fix_brief automaticallypip install stobo-mcpNo install with uvx:
uvx stobo-mcpAdd to your Claude Desktop config.
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"stobo": {
"command": "stobo-mcp",
"env": {
"STOBO_API_KEY": "your-api-key",
"STOBO_BASE_URL": "https://api.trystobo.com"
}
}
}
}STOBO_API_KEY is optional. Every audit and fix brief tool works without one. API key support for premium tools is coming soon.
Full setup guide at docs.trystobo.com.
pip install --upgrade stobo-mcpRun check_connection. Claude tells you if a newer version is available.
generate_fix_brief: structured fix brief from any completed audit. Free, instant, no LLM.audit_site or audit_article run.check_connection reports when a newer version is available.MIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.