"author:asselwan" matching MCP servers:

Verifies AI agent actions by exercising public web surfaces and returning evidence-based pass/fail verdicts on real outcomes, rather than trusting self-reported success messages.