Anthropic's MCP Goes GA: Agent Commerce Infrastructure Arrives
Anthropic's Model Context Protocol reached general availability with a wave of enterprise integrations, making structured server exposure the new baseline for AI discoverability. Businesses without MCP endpoints or llms.txt files are now actively invisible to the fastest-growing class of AI agents.
This Week's Signals
MCP General Availability Triggers Enterprise Integration Wave Across SaaS Platforms
Why it matters for your score
With MCP out of beta, Claude agents can now query, authenticate, and transact with MCP-enabled services natively — no custom integration required. Businesses that have published an MCP server endpoint are surfaced as actionable options inside Claude; those without one are skipped entirely. This is the agent-commerce equivalent of not having a mobile-friendly website in 2015.
Perplexity Expands 'Shop with AI' to Service Businesses, Cites Structured Pricing Data as Primary Ranking Signal
Why it matters for your score
Perplexity's commerce layer now covers service categories — agencies, consultants, SaaS — not just physical products. Their internal documentation confirms that machine-readable pricing pages and schema-tagged service offerings directly influence which businesses get surfaced in answer cards. If your pricing lives only in a PDF or behind a 'contact us' wall, Perplexity's agent cannot recommend you.
llms.txt Spec Crosses 40,000 Verified Deployments as Google Confirms Indexing Crawler Reads the File
Why it matters for your score
The llms.txt community tracker crossed 40,000 confirmed deployments, and Google's Search Central documentation quietly confirmed that Googlebot's AI-mode pipeline now parses llms.txt alongside robots.txt to determine which content to surface in AI Overviews. This is the first confirmed signal that llms.txt influences Google's own AI surfaces — not just third-party agents — making deployment urgent for any business chasing AI-native search visibility.
What to do this week
- 1
Deploy or audit your llms.txt file this week — Google's confirmed parsing means this is no longer optional for AI Overview visibility. Use Faro's /tools/llms-txt generator to produce a spec-compliant file with accurate service descriptions, and verify it's live at yourdomain.com/llms.txt before the next crawl cycle.
- 2
Add machine-readable pricing to at least one core service page using PriceSpecification schema, then validate it with Faro's /tools/ai-schema tool. Perplexity's ranking signal confirmation means unstructured pricing pages are actively costing you recommendations in the fastest-growing AI commerce surface right now.
- 3
Run Faro's /tools/ai-readiness-scan to check whether your domain is blocking MCP-relevant crawlers in robots.txt and whether you have an MCP server endpoint listed. With MCP now GA, the gap between businesses with structured agent endpoints and those without is compounding weekly — the scan will show you exactly where you're leaking AI referrals.