Composability
Definition
The design principle that agent systems should be built from modular, interchangeable components that can be combined in different ways to solve different problems. A composable agent system separates: the model (reasoning engine), the runtime (execution environment), the tools (actions), the memory (state), and the orchestration (coordination). Each component can be swapped, upgraded, or scaled independently. Composability is what enables the agent ecosystem — standardized interfaces (MCP, A2A) allow components from different providers to work together.
Builder Context
Design your agent system for composability from the start. Practical steps: (1) use standard interfaces for tool integration (MCP over custom APIs); (2) separate your agent logic from your model provider (don't hardcode OpenAI-specific features); (3) externalize state and memory (don't store important state only in the conversation context); (4) make each component independently testable and replaceable. The most common composability failure: coupling business logic to a specific model or framework, making it painful to switch when better options arrive.