B2B SaaS with no signal of agent-traffic differentiation (Signature-Agent header, agent-access policy)
This check looks for whether your site can tell the difference between human visitors and AI agents, for example by parsing the 'Signature-Agent' header (a piece of information some AI browsers send to identify themselves) or having a stated agent-access policy. This B2B SaaS site showed no such signal. Tools like OpenAI's Atlas and Operator use a standard Chrome browser identity, so the Signature-Agent header is currently the main way to tell an AI agent apart from a human visitor. Without that distinction, agents and human customers hit the exact same rate limits and pricing, so you can't price, throttle, or personalise for agent traffic as agent-driven commerce grows. Your engineering team should detect the Signature-Agent header in your edge layer (the part of your infrastructure that handles requests first) and apply different rate limits, pricing tiers, or content surfaces to agent traffic. Consider an agent-only API tier if agent usage grows. Budget around a week to build and test this.
Why this matters
Tools like OpenAI's Atlas and Operator use a standard Chrome browser identity, so the Signature-Agent header is currently the main way to tell an AI agent apart from a human visitor. Without that distinction, agents and human customers hit the exact same rate limits and pricing, so you can't price, throttle, or personalise for agent traffic as agent-driven commerce grows.
How to fix it
Your engineering team should detect the Signature-Agent header in your edge layer (the part of your infrastructure that handles requests first) and apply different rate limits, pricing tiers, or content surfaces to agent traffic. Consider an agent-only API tier if agent usage grows. Budget around a week to build and test this.