Blockworks launched Agentic Detection on September 17 as a feature within its Monitoring product, using proprietary AI agents to issue crypto risk and activity alerts in roughly five seconds. The previous manual analytical review took eight to 10 minutes, representing a reduction of about 99%. The system monitors X, Discord, Telegram, GitHub, governance forums and more than 1,000 news sources. It covers over 1,000 assets across 55 subcategories, including smart contract exploits, governance proposals, bridge pauses and abrupt changes in project activity. Alerts are initially marked Agent-detected until a human analyst verifies them, giving institutional users speed while distinguishing machine-generated signals from reviewed findings. Alerts can be delivered through Slack, Telegram, email, webhooks and APIs. Blockworks is building the feature on Messari data and research infrastructure acquired in June for more than $10 million. Messari's Monitoring product had served crypto exchanges, risk-management desks and token-listing teams. Existing topic information described an earlier or differing deployment as covering more than 500 assets and over 60 event types, with roughly 40 terabytes of data, seven years of analyst coverage and a June 12, 2026 acquisition completion date; those details differ from the latest account and are retained as historical or conflicting specifications.