Triggers

The runtime continuously observes chains, protocols, and data sources for changes that may require the agent to act. These changes can include market movements, protocol events, scheduled intervals, or user-defined conditions. When the runtime detects that something meaningful has occurred, it sends a trigger to the BunnyOS Framework. This trigger signals the agent to run its logic, evaluate the current situation, and determine the next appropriate step.

By separating decision timing from decision logic, the runtime allows agents to remain efficient and responsive. The runtime decides when the agent should think, not what the agent should decide. This structure ensures that agents do not run unnecessarily, but also never miss important conditions. It creates a stable and predictable execution rhythm that keeps agents synchronized with real-time market and protocol activity.