I built an x402 buyer-intelligence agent that turns one B2B offer into 3 source-backed potential buyers

I’m experimenting with a simple agentic workflow called Revenue Scout. The input is intentionally small: you describe what you sell. From there, the agent: interprets the offer and infers an initial ICP researches potential B2B buyer companies using public sources requires evidence for account fit rather than returning generic company lists ranks the strongest candidates returns exactly 3 potential buyers explains why each company may be relevant includes source-backed evidence and confidence signals explicitly marks things like buying authority, budget, demand, or timing as UNKNOWN when they cannot be verified One design decision I’m testing is whether reducing the input burden actually makes the agent more useful. The user shouldn’t have to manually provide an ICP, target-account list, or research plan before the agent can start working. The current Tier A call costs 0.25 USDC through x402. I’m not trying to prove that these are sales-qualified leads. The output is buyer intelligence: a researched shortlist that another agent or human can use as the next step in a sales workflow. Right now I’m specifically trying to validate one question: Would agent builders or B2B service providers actually pay for this output rather than doing the research themselves? I’d especially appreciate feedback from people building sales agents, marketplace agents, or autonomous commerce workflows.