Ready-made analyses
Each prompt is a slash command that runs a fixed set of tools.Questions by task
Visibility and competitors
- “How visible is this brand over the last 28 days, and against whom?”
get_overview - “Which competitor gained the most share of voice on Gemini?”
list_competitors - “Which brands do the answers name that we do not monitor?”
list_competitors
Prompts
- “Which prompts name a competitor and never this brand?”
list_prompts - “Which prompts name no company at all, and what should replace them?”
list_prompts,list_suggestions - “Show me every answer to this prompt on Perplexity.”
list_prompt_runs,get_answer
Sources and searches
- “Which domains do the answers cite most, and where does our own site rank?”
list_sources - “Which of our pages do the answers cite, and which do they read without citing?”
list_own_pages - “Which web searches does ChatGPT run before it answers, and which of them name a rival?”
list_searches
Actions
- “What are the three most worthwhile things to do this week, with the evidence?”
list_actions - “Which pages cite our competitors and never us?”
list_actions
Perception and shopping
- “What objections do the platforms raise about this brand, and since when?”
get_perception - “Which products do ChatGPT’s product cards show for our prompts, and at what price?”
list_products
The website
- “Do the AI crawlers reach the site, and which pages fail for them?”
get_crawler_visits,get_crawlability - “Which landing pages get visits from AI assistants, and do the answers cite them?”
get_ai_visits
Several projects
- “Compare the visibility of every client this month.”
list_projects, thenget_overviewfor each project
Tips
- Name the period. “Over the last 28 days” gives a figure that you can compare with the dashboard.
- Ask for the answer behind a claim.
get_answerreturns the answer word for word, so the assistant can quote it. - Keep wide periods for when you need them. A call over a wide period costs more of the rate limit. Plan limits states how much.