---
name: geo-visibility-audit
description: Audit any domain across all 9 AI answer engines (ChatGPT Search, Perplexity, Claude, Google AI Overviews, Google AI Mode, Gemini, Grok, DeepSeek, Copilot), verify edge WAF bot-spoofing behavior, and measure non-branded citation Share-of-Voice.
---

# Argosor 9-Engine GEO Visibility Audit Skill

Use this skill whenever the user asks to check how their brand or domain ranks in AI search engines (ChatGPT, Perplexity, Claude, Google AI Overviews, AI Mode, Gemini, Grok, DeepSeek, Copilot) or why AI crawlers are not citing them.

## Step 1 — Run the 9-Engine GEO Audit & WAF Bot-Spoofing Probe
Call the `geo_llm_visibility_audit` MCP tool (or `POST https://argosor.com/v1/geo/audit`):
```json
{
  "brandName": "Allometrik",
  "domain": "allometrik.com",
  "promptTopic": "autonomous mcp infrastructure consulting",
  "competitors": ["llmpulse.ai", "dataforseo.com"],
  "llmEngines": [
    "chatgpt_search",
    "perplexity",
    "claude",
    "google_ai_overviews",
    "google_ai_mode",
    "gemini",
    "grok",
    "deepseek",
    "copilot"
  ]
}
```

## Step 2 — Check Edge WAF Bot-Spoofing & Agent Interfaces
Inspect `wafBotSpoofingProbe` and `agentInterfacesAudit` in the response:
- Verify that `GPTBot`, `ClaudeBot`, and `PerplexityBot` receive `200_ok_reachable` at the edge (no false 403 WAF blocks).
- Verify that `/llms.txt` is not returning a false-200 HTML SPA shell (`spaFalse200: false`).
- Check for `Content-Signal: search=yes, ai-input=yes` in `/robots.txt`.

## Step 3 — Audit 3-Tier AI Crawler Taxonomy & Analytics Attribution
Call `agent_crawler_traffic_attribution` to separate Training Crawlers from AI Search Indexers and Live User-Triggered Agent Browsers, and output ready-to-paste GA4, Plausible, and PostHog regex filters.