📘 This article introduces the Botify MCP, which brings your Botify data into your own LLM. Botify MCP is available with any Botify plan.
Overview
Every prompt below is a starting point rather than a fixed command. Your assistant understands follow-up questions, so you can narrow a result by folder, date range, or bot name without starting over, and you can rewrite any prompt in your own site's terms.
Square brackets identify the parts to replace with your own values, such as [project], [page], or [query].
Prompts by task
Crawl budget and page health
Which pages returned a 4xx or 5xx status code in the last crawl, and do any of them still have internal links pointing to them?
Structured data across a template
Do any product pages in the /shop/ folder have missing Product schema?
Define your expected field values before the run, or there is nothing to validate against.
Title, description, and H1 changes before they ship
Preview what this page's title tag would look like if it started with our brand name.
Pull the performance baseline in the same pass as the preview. A change shipped without a recorded before state cannot be evaluated afterward.
Search engine crawler and AI crawler activity
Compare Googlebot and GPTBot hit volume over the last 30 days, and flag any spike in 403 or 404 responses.
For the full walkthrough, see Botify MCP Use Case: Separating Search Engine and AI Crawler Traffic.
AI crawler access and brand citations
Are GPTBot and PerplexityBot able to reach our site right now, and does our brand come up if I ask [prompt] on ChatGPT and Perplexity?
For the full walkthrough, see Botify MCP Use Case: Checking AI Crawler Access and Brand Citations.
Search Console performance and ranking decay
Show week-over-week changes in clicks, impressions, and average position for our top 20 pages by traffic.
Whether a ranking or traffic shift lines up with a change you made
Did rankings for [page] change after the release on [date]?
What to write next, and whether a page already exists
Cluster these keywords by intent, and tell me which ones we do not have a page for yet.
Keep the generated question set as a reusable reference rather than regenerating it for every brief. The same clusters resurface across topics.
What search results and AI answers show right now
What does the search engine results page (SERP) look like for [query] right now, including any AI Overview or shopping results?
Your brand's entity and search presence
What does Google's Knowledge Graph show for our brand, and what comes up if I search for us directly?
Most valuable for catching wrong attributions rather than confirming right ones. A clean result is unremarkable; the finding is when something is off.
A weekly report across crawl, Search Console, and analytics
Pull crawl status, Search Console clicks and impressions, and GA4 sessions for [project] for the last 7 days.
For the full walkthrough, see Botify MCP Use Case: Building a Weekly Search Report.
Page performance answers without an export
Which product pages are missing meta descriptions, and how are they performing in Search Console?
For the full walkthrough, see Botify MCP Use Case: Self-Serve Page Data for Content Teams.
Your own business data joined to search data
Load this revenue CSV and join it against our crawl data by URL.
For the full walkthrough, see Botify MCP Use Case: Joining Your Revenue Data to Search Data.
Whether an extraction is trustworthy before you build on it
Check the extracted price and availability fields against what is actually on the page for these 50 URLs.
Page load time performance
Investigate the reasons why load times for the pages crawled during the latest crawl of myproject-org/myproject-brand.com have deteriorated over time.
Read next
Related topics