Skills · MCP · Claude API

AI Safety Hub

Ask SafetyBot about Anthropic's safety research, the Claude Constitution, the Opus 4.6 Sabotage Report, or responsible AI deployment in financial markets.

🛡️
SafetyBot
Online — powered by Claude
Welcome to the AI Safety Hub. I can answer questions about Anthropic's safety research — the Claude Constitution, the Opus 4.6 Sabotage Risk Report, the Responsible Scaling Policy — as well as how these principles apply to responsible AI deployment in financial markets.

Under the hood, I use a Skills pattern (knowledge documents loaded from S3) and an MCP-style tool (I choose which documents to load based on your question). Ask me anything.
SafetyBot · just now
AI-generated responses · Not financial advice · Powered by Claude Sonnet 4.5