: Use a "before_agent" method to intercept user requests or an "after_agent" method to scan model responses before they are delivered.
: The blocking logic should be decoupled from the primary agent. This allows you to update security policies or "constitutions" without having to retrain or reconfigure the main task-oriented agent. Step-by-Step Development Process blocking agent
: The blocking agent needs access to the current "state" (conversation history) to identify context-specific risks that might not be apparent in a single message. : Use a "before_agent" method to intercept user
: Explicitly list what the agent is not allowed to do. This might include blocking the output of API keys, preventing the execution of destructive commands (like rm -rf ), or filtering toxic language. Step-by-Step Development Process : The blocking agent needs
: When a block occurs, the system must handle it gracefully—such as providing a standardized "I cannot fulfill this request" response—rather than just crashing or failing silently. Key Patterns in Modern Agentic Systems How to Build Reliable AI Agents (without the hype)
: This is the "brain" that analyzes incoming data against your rules. In production systems, this often involves a smaller, faster model (like GPT-4o-mini or Claude Haiku) optimized specifically for classification and risk detection.
To develop a detailed piece, you must integrate several foundational building blocks: