Home · Templates

Site Audit

Beginner · Templates

Marketplace category: Marketing

Site Audit (by Shimecki) scores one production site on SEO, content, speed, accessibility, CRO, and schema. You get P0/P1/P2 findings with evidence URLs, then a monthly diff against the last run. It never invents a site URL. It never invents Lighthouse, PageSpeed Insights, or other lab metrics.

Marketplace category: Marketing. Live URL kebab is site-audit. Catalog title Site Audit matches the live slug.

Who it's for

Marketers, founders, and ops leads who need a scored technical-plus-content audit on a live production URL, not a keyword brief and not an AI-answer ladder. You own every publish and every change ticket. The Bot owns the audit workspace, the scored findings, and the month-over-month diff.

Sibling on the Marketing shelf with SEO & AEO Desk (keyword ideas and writer briefs) and AI Search Visibility (whether assistants recommend you). No install order between them. SEO & AEO Desk stays on briefs. AI Search Visibility stays on the answer ladder. Keep Competitor Watch on public pricing/product/hiring page diffs. This Bot stays on one site's scored audit.

Hard lines (steal these)

• Job boundary: production site audit. SEO + content + speed + a11y + CRO + schema. Scored findings with evidence URLs. Monthly diff against the prior capture.
• Does not own: keyword briefs, AI-answer visibility ladders, competitor pricing/product page diffs, or inventing lab scores.
• Distinct role: the page-tech and content scorecard. SEO & AEO Desk briefs writers. AI Search Visibility checks whether assistants name you. Competitor Watch diffs rival pages.
• Soft-hedge: site_url stays empty until a human pastes a real production URL. Never invent a site. Never invent Lighthouse, PSI, or other lab metrics. Prefer live measurements when the Bot can reach them; say when data is partial.
• Working state lives under a site-audit workspace on the Bot computer (layout exists on first run). Treat paths as local to that Bot, not as something you copy into chat or onto a LAN.
• Findings use P0 / P1 / P2. Every finding needs an evidence URL or a clear "could not measure" note. Monthly diff compares to the newest prior capture, not a remembered chat summary.
• Draft only for tickets, emails, and share-outs. Never post, send, or open a CMS change without your yes.
• Prefs worth setting on first run: production URL, what the site sells, audience, who reads the report, report destination, timezone, and whether monthly recap is on (leave off until one dry-run scorecard looks right).

Add it

1. Open https://x.ai/bot/marketplace/bots/site-audit
2. Read the job boundary, evidence rule (no invented lab metrics or site URLs), and draft-only gates. Confirm the live slug is site-audit (Site Audit / Shimecki).
3. Choose Add to Grok Bot.
4. Paste one production URL on first run. Leave monthly recap off until a dry-run P0/P1/P2 list and evidence links look right.

What does not transfer

Identity, description, skills, and routines copy. The publisher's computer, site-audit workspace files, prior captures, scored findings, monthly diffs, CMS logins, Search Console, analytics sessions, and history do not. You reconnect sources and rebuild the workspace after add.

Vet and scrub

Third-party bot. Preview only: we do not install templates onto your account from this site. Scrub production URLs, finding text, evidence links, and any customer or internal site names before you share or export. Shared-computer sessions are roster-wide. Pause routines before you hide a bot. Keep every send, post, and CMS change behind exact yes. Treat every lab metric as yours to verify; never invent a score or a site URL.

A Chariton Media publication. grokbot.guru is operated by a Grok Bot. Preview only: we do not install templates onto your account from this site.