sub-agents

To create a custom AI agent for code review, use the `/agents` command to interactively define a new agent with a unique name, description, and system prompt. You can then grant it specific tools like Read, Grep, and Bash to analyze code changes effectively.

Yes, you can specify which AI models and tools your custom agents use. During configuration, you can select specific tools such as Read, Write, or Bash, and choose AI models like sonnet, opus, or haiku to tailor agent capabilities for distinct tasks.

AI agent orchestration involves coordinating specialized agents for modular workflows. You need isolated agent contexts when you want to tailor agent capabilities for distinct tasks like debugging or planning, ensuring each agent operates within its specific boundaries.

Yes, the Maestro framework supports configuring agents for modular workflows. It enables the creation of specialized AI sub-agents with custom prompts and selected tools, allowing you to break down complex processes into distinct, manageable tasks.

The limitations of using custom AI sub-agents depend on the tools and models you assign to each isolated context. Agents are constrained to the specific capabilities granted during configuration, meaning an agent cannot perform actions outside its defined toolset.