Cross-examine text for logical fallacies and manipulation — from the CLI, your browser, an MCP tool, or a Discord/Slack bot, with any LLM. It detects 57 techniques across fallacies, rhetoric, bias, and scam signals.