Anthropic MCP Goes GA as llms.txt Crosses 50K Domains
Model Context Protocol reached general availability this month with major platform integrations, while llms.txt adoption crossed a verifiable milestone — both signals that AI agent discoverability is shifting from experiment to infrastructure. Businesses without structured machine-readable context are now measurably less likely to surface in agentic workflows.
This Week's Signals
Anthropic MCP Hits General Availability With Shopify, Stripe, and Cloudflare Integrations
Why it matters for your score
MCP moving to GA means Claude-powered agents can now natively query business data, product catalogs, and transaction systems through standardized connectors — not just read static pages. If your business runs on Shopify or Stripe and hasn't published an MCP-compatible endpoint, you are already invisible to a growing class of purchasing agents. Faro's /tools/ai-readiness-scan now checks for MCP endpoint exposure as a scored signal.
llms.txt Adoption Surpasses 50,000 Indexed Domains as Perplexity Confirms It Uses the File for Source Ranking
Why it matters for your score
Perplexity's confirmation that llms.txt influences which sources get cited in answers is the clearest signal yet that the file has moved from courtesy to competitive advantage. At 50K domains, adoption is still sparse enough that publishing now is a differentiator — but the window is closing. Use Faro's /tools/llms-txt generator to produce a correctly formatted file with your business context, services, and preferred citation language in under five minutes.
Google's Structured Data Documentation Updated to Include AI Overview Citation Eligibility Signals
Why it matters for your score
Google quietly updated its structured data guidance to note that complete Schema.org markup — specifically LocalBusiness, Product, FAQPage, and Review types — is evaluated when determining eligibility for AI Overview citations. This is the first time Google has explicitly linked JSON-LD completeness to AI surface inclusion, not just traditional rich results. Incomplete or missing schema now carries a dual penalty: losing both standard SERP features and AI citation slots.
What to do this week
- 1
Audit your MCP exposure now — GA status means agents are actively querying MCP endpoints in production. Run Faro's /tools/ai-readiness-scan to see whether your business registers as MCP-discoverable, and follow the flagged recommendations to publish a minimal agent-facing endpoint if you're on Shopify, Stripe, or a supported platform.
- 2
Publish or refresh your llms.txt file this week using Faro's /tools/llms-txt generator. Given Perplexity's confirmed use of the file for source ranking, prioritize including your canonical service descriptions, geographic service areas, and a clear preferred-citation sentence — these are the fields most likely to influence how AI answers describe your business.
- 3
Run a JSON-LD audit against Google's updated AI Overview eligibility signals using Faro's /tools/ai-schema tool. Focus on filling gaps in LocalBusiness, FAQPage, and Review schema — incomplete markup now costs you placement in both traditional rich results and AI-generated answer surfaces simultaneously.